APSC 459 Report
Note to Pavle. Page 27, 28, 29.
T ABLE OF C ONTENTS
T ABLE OF C ONTENTS
1.1.2 Current Rubidium Gas Transition Locking System ........................................ 3
T ABLE OF C ONTENTS
1 | I NTRODUCTION
1.0
I NTRODUCTION
Laser cooling is a recently developed technique used to slow down atoms and subsequently cool down their temperature. This technique can be exploited to study and analyze the atom at a quantum level by manipulating ultra cold atoms.
Laser cooling of atoms, however, has proven to be a difficult task as it requires multiple frequency-variable lasers locked at very specific frequencies relative to one another. It is not sufficient to simply tune two lasers to appropriate wavelengths as the lasers will not maintain that output frequency. Lasers have a tendency to ‘drift’, or in more common terms, slowly shift their outputted wavelength. The current frequency locking system employed such as the one being used in the Quantum Degenerate Gas Lab requires expensive and complicated assembly procedure.
Single mode laser diodes commercially introduced in the last decade are acceptable in performing high resolution spectroscopy on atoms and molecules. However the tunability of these diodes is not sufficient for applications in laser cooling where the diode laser needs to lase at 780 nm. [3] The performance of the frequency tuning on these diode lasers can, however, be improved by introducing an external cavity that provides optical feedback to generate a more sensitive tuning.
The project is aimed to design and construct an external cavity diode laser (ECDL) that will tune its own frequency according to a frequency from another laser using diffraction grating; a Frequency Locking System of Grating Stabilized Diode Laser (GSDL). The approach in designing this frequency-tunable laser is through an external cavity consisting of a diffraction grating and an angularly adjustable mirror to send the diffracted 1 st order beam back into the laser for feedback known as the Littman configuration.
1
1 | I NTRODUCTION
1.1
Background
Although the focus of this report and project is centered on the design and construction of a grating-stabilized frequency locking system, background on laser cooling is important to recognize the significance of the frequency locking system and its purpose. This section briefly explains what laser cooling is and how it works as well as the existing technique employed to lock a frequency of a laser in the Quantum Degenerate Gases Laboratory
(QDG Lab) at the University of British Columbia.
1.1.1
Laser Cooling
D. J. Wineland first demonstrated laser cooling in 1978. It is essentially a method that utilizes lasers to freeze atoms in a contained space. The basic concept of laser cooling is to transfer energy and momentum from the photon to the atom in order to slow it down. The temperature of a collection of gas molecules is related to their kinetic energy (i.e. their average velocity). By slowing down the atoms in a localized region, it is possible to rapidly cool down the temperature of the gas to a fraction of a millikelvin. The following diagram will aid in the discussion.
Figure 1: One-dimensional simplified schematic of photon-matter interaction.
To simplify the discussion of laser cooling, we will only focus on the one-dimensional case. When a photon, propagating at a resonant frequency, strikes an atom, it is absorbed and an electron within the atom makes a transition to a higher state given by:
The atom in a random direction then spontaneously emits this energy.
However, if the atom is moving in the opposite direction of incoming light, it will see a higher frequency due to the Doppler shift effect. Therefore, by adjusting the frequency (ν p
) of the incoming photon slightly below resonance, the moving atom will see a resonant
2
1 | I NTRODUCTION frequency photon. The atom will absorb this photon, again causing a transition jump of the electron and then later a spontaneous emission. Looking closely at the energy components, the energy of the absorbed photon is:
The energy released by the atom upon spontaneous emission is given by (1). Since the resonant frequency ν
0
is larger than the photon frequency ν p
, the atom will lose more energy than it receives causing it to slow down.
When the atom travels in the opposite direction, it sees the photon frequency below ν p
and is therefore unaffected by its presence. In order to stop an atom traveling to the right, a light source is set up to the right of the atom with the incoming light pointed to the left.
Thus the use of the two lasers aimed in opposite directions will slow down the atom almost to a complete stop.
Expanding the design to a three dimensional environment requires three coupled lasers in the three respective coordinates of space. The atoms cannot be completely stopped since there is recoil when it spontaneously emits a photon in a random direction.
One of the setbacks of producing such an experiment is that light has to be output by a laser with frequency which is at a constant difference from an atomic transition. Rubidium, common gas that is cooled in such experiments, is going to be cooled in Dr. Madison’s lab.
The frequency offset difference required to cool Rubidium is 6.8 GHz. The current technique used to establish the required frequency offset is explained below. Figure 2 will aid in the discussion.
1.1.2
Current Rubidium Gas Transition Locking System
Figure 2: A similar setup to the current use of Rubidium vapor atomic transition frequency to stabilize diode laser frequency [2]
3
1 | I NTRODUCTION
The current FLS is able to lock its frequency to an atomic transition of Rubidium, and then this frequency is shifted by an Acousto-Optic Modulator (AOM). Specifically, the laser is back reflected from the rubidium using a glass wedge. The back-reflected beam is sent into a photodiode and then sent into a lock-in mechanism to determine the locking. This information is sent into the feedback to control the piezoelectric transducer to control the laser frequency. Now that the frequency of the laser is set at an atomic transition it has to be shifted to the correct offset. This is accomplished by the use of the AOM and lens system. Though the use of the atomic transition frequency locking has advantages in accuracy, it requires expensive components. In the application of laser cooling, this is not an ideal setup since six laser systems are required and each setup like the one shown in
Figure 2 costs above $5,000.
With the development of high-speed photodiode detectors that have a capability of detecting signals up to 10 GHz, a more elegant and much cheaper setup can be implemented to replace the existing FLS. The system that we propose to develop will be simpler and more cost effective than the existing one. The design will force one laser (the
Slave Laser) to output a specific frequency such that the frequency between the two lasers is constant. The complete setup of the system is shown below along with the description of how it works.
4
1 | I NTRODUCTION
1.2
Project Objective
1. Design modification of GSLD
The current design utilized by Dr. Jones and Dr. Madison are intended for more rugged, all around purposes and as such contain extra components that complicate and interfere with the purposes of this project. A simpler more dependable design is required that satisfies the purposes of only this project.
To address this we intend to submit a final design for approval by Dr. Madison and
Dr. Jones by January 20, 2007.
2. Construction of 1 st GSLD a.
Laser Temperature Stabilization
Using a TEC cooler system, this portion of the design will likely consist of a thermistor, thermal paste, a TEC cooler and an off the shelf electronic control unit
(TEC 2000 2A / 12 W). b.
Controlled Angle of Diffraction Grating
The angle of the diffraction grating will be controlled electronically via a piezo unit.
The specifics of the construction vary depending on which GSDL configuration is constructed. The two designs however do share several components such as the laser diode, the fiber coupling optics, a mirror, an adjustable stage that is driven by a piezo unit pushing against a thin brass structure and of course the diffraction grating.
However the angles and dimensions depend entirely on the configuration decided upon.
The controlling of the PTZ is necessary in both configurations and will be done by the circuitry provided by the electronics shop. The voltage levels applied to the PTZ driver will vary between –15V and + 15V. These voltages will then be amplified between –150V and 150V. The wiring of the PTZ driver can be found in Appendix
B. c.
Operational Laser Diode
The laser diode is not to be designed but is to be incorporated into the GSLD design.
Dr. Jones and Dr. Madison have already obtained the Laser Diode that is to be utilized.
We intend to have a completed GSLD ready for testing by February 16, 2007.
4. Combination of System Components
The outputs of the Master and Slave Lasers are to be combined with the PI Servo and frequency comparing electronics. Combining all the pieces of the system is not expected to be difficult with the help of the sponsors; however tuning the PI Servo electronics to optimize the locking system may prove tedious.
5
1 | I NTRODUCTION
The entire system should be assembled by March 16,2007 at which time the FLS can be tested in a full system test. This requires that all of the hardware be operational and connected leaving only the tuning of the PI Servo to complete the system.
5. Operational Frequency Locking System
As mentioned, the tuning of the system may prove the most time consuming towards the final stages of the project. If an acceptable performance is obtained (the tuning will not be overly critical as the lasers will be temperature stabilized to deter against shifts in outputted frequency) with time to spare, deliverables in the actual laser cooling experiment may also be assumed.
The last and final deliverable is the entire operational system and the delivery date is scheduled for March 30, 2007 to allow time for us to create the final report.
6
1 | I NTRODUCTION
1.3
Scope and Limitations
The scope of this report is to analyze the design specifications of the Frequency Locking
Systems as well as the results of testing procedures used during the design and construction. Also, details from the testing of the FLS are identified in this document.
Specifications regarding the design and schematics of external electronics including photo-detector, frequency dividers, digital comparators, and the PI servo control, however, will be briefed but not closely examined as the core focus of the project lies in the design and construction of the grating stabilized diode laser. The omission of such details should not produce a large impact on this report as the electronics provided had no major changes and issues with the progress of the project.
1.4
Organizations
This report is composed of sections describing the theory of frequency locking to illustrate the different approaches in designing a frequency tunable laser; Littman and Litrow configuration. The discussion section will also include an analysis of the initial design, final design, temporary tabletop setup, and changes made to the designs. A general overview of the complete negative feedback system of the FLS is discussed under discussion section.
7
2 | D ISCUSSION
2.0
D ISCUSSION
This section of the report is separated into theory, method, design, data, results, and laser locking. Before the design of the grating stabilized diode laser, different setups of external cavity feedback systems were analyzed; particularly the Littman and the Littrow ECDL setups. These are the focused components of the complete laser frequency locking system that is developed within this project. Different analysis and testing procedures were used during the design of the ECDL and these are closely examined to present the rationale on the formation of the initial design as well as modifications before and after the construction of the ECDL. Finally the results from the finalized GSDL is presented and evaluated along with the stability and efficiency in the laser locking capability of the complete system.
8
2 | D ISCUSSION Theory
2.1
Theory
Prior to the commencement of the project, two setups of the GSDL were considered and discussed in meetings with the project sponsors. These two different setups of GSDL are the Littrow and Littman configurations. Both of these setups involve an external cavity diffraction grating that splits the single mode laser into a visible 0 th order and a 1 st order beam. The 1 st order beam is then beamed back into the diode as a source of feedback.
What differs between the two setups is the way the 1 st order beam is sent back into the diode as discussed in the following sections under this Theory section.
Engineering an ECDL with mode-hop free tuning over 2+ GHz is a challenge in mechanical precision and accuracy. The goal in this project was to obtain a large enough mode-hop free bandwidth to cover a span of 8 GHz from the master laser. This, ideally, is approximately the width of the Rubidium Absorption Spectrum of the Rubidium-85 and
Rubidium-87 isotope gases.
2.1.1
Rubidium Electron Transition
Since the current frequency locking system is locked to Rubidium transition frequency and this frequency needs to be locked by the GSDL, it is necessary for the GSDL to be able to obtain a bandwidth of frequency around that region. Rubidium spectroscopy has a region of interest at 780 ± 2 nm. This wavelength in frequency is approximately 3.846×10 14 Hz
(or 384.6 THz). The rubidium absorption scan shown in figure below displays 4 minimums representing the four absorption peaks of both Rubidium-85 and 87 atoms.
9
2 | D ISCUSSION Theory
As can be seen from the figure, the four absorption minimums cover an approximate width of a little over 8 GHz. One of the secondary goals in this project is to attempt 8 GHz of bandwidth.
2.1.2
Grating Stabilized Diode Laser
As discussed previously, the Slave Laser needs to be able to constantly tune its frequency in order to output a frequency difference of 6.8 GHz. This fine-tuning can be established by designing a laser with an external cavity, diffraction grating and a mirror. The diffraction grating is used to reflect light onto a mirror, and back into the laser diode. By adjusting the length of the cavity or the angle of either the mirror or grating (depending on the configuration used), the laser is able to select the frequency that it will amplify.
There are several existing designs for GSLDs. The two most common designs are referred to as the “Littrow” and the “Littman” configurations. One of these configurations will be employed by this project. Which configuration is to be constructed will be based on whether or not a working Littman model can be constructed before January 2007.
The Littrow design uses the 0 th order diffraction beam to select the frequency of amplification. This design however changes the angle of the output beam and therefore the coupling optics must also physically move in sync with the GSDL components in order to properly couple the outgoing light. If this complexity can be avoided it would greatly reduce the assembly time of the GSDL.
10
2 | D ISCUSSION Theory
2.1.2.1
Littrow Configuration
Figure 5: The Littrow Configuration: 1) Laser Diode Output 2) Amplification Beam
3) Diffraction Beam 4) GSDL Output
The Littman design may however prove unacceptable as the amplifying beam that returns into the laser diode is of a higher order than in the Littrow design and must be reflected twice before returning to the laser diode. This greatly reduces the intensity of the light and in turn the amplification and selection of the desired frequency from the laser diode. A setup of the Littman configuration is being assembled and will be tested by our team before the January term. If it is shown that the Littman configuration can lase, it will be constructed. Otherwise, the Littrow design will have to be carried out.
2.1.2.2
Littman Configuration
Figure 4: The Littman Configuration: 1) Laser Diode Output 2) 0 th Order Diffraction Beam
3) 1 st Order Diffraction Beam 4) Diffracted Amplification Beam
The Littman configuration, as shown above, is able to select a particular frequency by adjusting the angle of the mirror. This is done by the use of a Piezoelectric Transducer
11
2 | D ISCUSSION Theory
(PTZ). The 1 st order diffraction beam bounces off the mirror and back into the diode laser.
This is how the lasing frequency is selected. The 0 th order diffraction beam is used as the output. The Littman configuration is a more desirable design as the GSDL output is always at a constant angle and is therefore easier to couple into a fiber.
2.1.3
The Frequency Locking System
Figure 3. Schematic Diagram of the Overall Project
From Figure 3, the Master Laser will be locked to a certain Rubidium atom transition and will be able to output light at a constant frequency. The light from the Master will be coupled into a beam splitter by the use of fiber optics. The Slave Laser will also direct a beam towards the beam splitter by the use of a mirror. The two signals will interfere and the resulting beatnote (ideally at a frequency of 6.8 GHz) will be detected by the highspeed photodiode which is a part of the Detector and Frequency Divider System (DFDS).
This signal will then be divided down by 64 by the two 1/8 frequency dividers. Division of the beatnote signal is carried out since the Frequency to Voltage Converter (FVC) only works up to a frequency of about 125 MHz.
The FVC system takes in two inputs, the divided beatnote signal and a reference signal provided by the Direct Digital Synthesizer (DDS). The DDS is essentially a frequency generator which will provide a reference sine wave with a frequency of 106.25 MHz (6.8
GHz/64). The FVC will then output a voltage which is proportional to the difference of the frequencies from the two signals (i.e.
ε
ν beat –
ν
DDS
). This voltage will be the error signal that will be used as part of the feedback loop to control the Slave Laser. When the
12
2 | D ISCUSSION Theory voltage of the error produced by the FVC goes to zero, it means that the Slave Laser has established a frequency offset of 6.8 GHz.
The error voltage produced by the FVC is fed to a PI Servo. The PI servo will then produce two signals which will be used to adjust the frequency output of the Slave Laser.
The low frequency signal produced by the PI Servo will change the angle of the mirror in the Slave Laser, while the high frequency signal will adjust the control current injected into the laser diode. Both of these effects will change the frequency of the Slave Laser and will force it to lase at a 6.8 GHz difference from the Master Laser.
2.1.4
The Negative Feedback Control System
The complete Frequency Locking setup can be simplified into a negative feedback control system as illustrated by a block diagram in figure below.
Reference
Signal
+
-
PI Controller
Master Laser
+
GSD Laser
Output
Frequency
The error signal generated from the reference with the output frequency is determined from a frequency difference between the master laser frequency and the slave laser frequency imposed on a synthesized signal. This frequency error is generated at large difference between the master and slave laser frequency. When the slave laser frequency is tuned fairly close to the master frequency (from GHz down to the MHz range), the error signal is switched from frequency difference to phase difference between the master and slave laser. This is done in order to match the two laser phases when the two frequencies are locked.
13
2 | D ISCUSSION
– Methods and Testing Protocols
2.2
Methods and Testing Protocols
Several optical components were used during the design, construction, and testing of the frequency locking laser. Most of the electrical equipments and optical components were provided without the need for purchase from the QDG lab. The electronics and optical devices used as well as their models and specifications are listed below. Most of these optical components are not part of the final design of the laser but are used in testing the frequency locking system to help debug and determine issues with slave laser stability.
2.2.1
Experiment Equipments and Parts
The following tables lists a set of electronic equipments that are used. Nothing needed to be purchased for this project. Only purchase made was a new Piezoelectric driver after the original broke during the project.
Electronic Equipments
Equipment
Current Controller
Temperature Controller
Power Meter
Infrared Viewer
Frequency Lock Box
Oscilloscope
Spectral Analyzer
Power Supply
Power Supply
PTZ Driver
Tabletop Components
Model
Thorlabs LDC 500 500 mA
Thorlabs TEC 2000 2A/12W
Newport: Power/Energy meter – Model 1825C
N/A
UBC-Physics & Astronomy E04-010
Tektronix
Tektronix RSA 3303 A Real-time Spec
Analyzer DC-3 GHz
BK Precision DC Regulated Power Supply
1627A
Lambda Dual Regulated Power Supply LPD
422 FM
Dual Low Voltage PTZ Driver YJ048-Q7
Component
Mirror Mount
Isolator
Price
NPR
NPR
Laser Diode
Diode Clamp
NPR
NPR
Piezoelectric Driver #1 NPR
Piezoelectric Driver #2
NP = No Purchase Required (Part of equipments available in QDG Lab)
14
1
1
1
2
Units
10+
1
2 | D ISCUSSION
– Methods and Testing Protocols
Developed Equipments
Component
Grating Holder
Grating Mount
Mirror Mount Stand
Base and Plates
2.2.2
Experimental Internal Setup of the Laser
Price
NPR
NPR
NPR
NPR
Units
1
1
1
1
The initial laser used for testing cavity lengths, angles, and positions was enclosed itself before beamed onto the diffraction grating. This enclosure isolates temperature change of the diode laser. The design is as shown.
2.2.3
Experimental Setup of the External Cavity
In order to design a portable enclosed laser box, angles needed to be determined to achieve maximum injection and largest bandwidth. Initially a few random angles, α, between 0 and
45 (between grating and the 0 th order beam in Figure below) degrees were selected to test the bandwidth. It was found that the larger the angle, the better the bandwidth of the laser.
Approximately 40 degrees was the maximum attainable angle since any angle beyond that would send the first order beam almost towards the diode allowing no space for the mirror to move or the first order beam would cross over to the other side of the laser. (might have to re-explain this part)
15
2 | D ISCUSSION
– Methods and Testing Protocols
The mirror is placed on a mirror mount with knobs that allow adjustment of vertical angle
(the knob on top) and horizontal angle (the knob on the bottom side). Adjusting the horizontal angle of this mirror mount changes the frequency of the laser. The first order is then sent toward a mirror that is only placed to minimize the lab area of this experimental setup. The beam from the mirror is then sent through an isolator which is then sent through a series of mirrors terminating in a fiber optic pickup from the spectrometer.
Mirrors after the isolator needed to be perfectly aligned such that the beam being picked up by the optic coupler receives maximum intensity. This is done by shooting a beam forward and backward in the setup above and matching beam locations on each mirror.
Isolating the beam from part A to part B prevents the beam used for coupling from shooting backwards into the laser diode.
Several diffraction grating and mirror mount angles were selected and tested to determine the allowed frequency before mode-hops using the spectrometer. These angles included 10,
20, 30, and 42 between the reflected 0 th order beam and the diffraction grating. Every change in angle required replacing the laser, diffraction grating, and mirror mount position as shown in the figure below.
16
2 | D ISCUSSION
– Methods and Testing Protocols
2.2.3.1
Bandwidth Optimization
This part is to describe the setup to achieve maximum bandwidth between mode-hops on the Rubidium absorption curve.
The experimental setup in determining the bandwidth using the spectrometer was not sufficient to verify whether the laser frequency was anywhere close to the desired master laser frequency of 780 nm (+/- a few nm) or
(something THz). Since the resolution of the spectrometer was relatively slow, it’s hard to determine whether mode-hopping occurred by sliding the frequency as the peak on the spectrometer jumped across the scan. In order to see if the laser was operating at the
Rubidium transition frequency without mode-hopping, the frequency would have to be linearly varied to generate a graph of frequency change. This frequency graph is referenced by Rubidium absorption curves in order to determine the proper frequency range the laser is able to achieve. The piezoelectric driver is ramped to oscillate the mirror slightly in angle such that a frequency of the laser is oscillated linearly. This laser is then sent through Rubidium gas cell and then received by a photodetector as shown in the figure below.
[Figure showing the setup to display Rubidium Absorption]
The beam input into the photo-detector varies as the frequency is oscillated since at certain frequencies, the Rubidium cell electrons undergo photon energy absorption, and emission simultaneously. This then allows a linear graph of Intensity vs. Frequency curve as shown in figure below.
17
2 | D ISCUSSION
– Methods and Testing Protocols
[Plot of Intensity vs Frequency Rubidium Absorption curve]
Finally, maximum bandwidth needed to be achieved and hence that meant the 1 st order beam would have to be tested along different positions of the back-reflecting mirror. This was accomplished by sliding the mirror mount and stand in a parallel fashion along the direction of the mirror. Unfortunately, this was very ineffective since sliding perfectly parallel is not very easy when a slight change in angle will stop the injection of the laser.
Which meant that every time the mirror mount was moved, the mirror mount knobs would have to be readjusted. To overcome this problem, a linear stage was used to hold the mirror mount such that the linear stage allows smooth sliding of mirror to adjust 1 st order beam position. Different points on the mirror were tested ranging from 5 mm to 30 mm
(from the opposite end of the pivot). With each position, data is recorded directly from the oscilloscope to measure amplitude of the injected signal. This data set is then graphed to display a relation between the location of 1 st order beam on the mirror and signal intensity.
(someone will have to read over this part, I have no idea what I’m talking about)
A temporary amplifier was used in the hopes of expanding the magnitude of the oscillation in which the piezoelectric driver was sweeping at. The amplifier schematic is displayed in the figure below. Expanding the oscillation ideally would expand the frequency range in order to maximize the bandwidth.
[Amplifier Schematic Figure]
18
2 | D ISCUSSION
– Design Proposal
2.3
Design Proposal
With the experimental setup completed and tested to optimize laser frequency, an initial design of a moveable enclosed laser is proposed with proper schematics. The exact schematics of each component are listed in the Appendix of this document.
2.3.1
Design Criteria and Requirements
Criteria’s were given in order to meet a more narrow design and specific proposal. The criteria’s are as follows:
The laser needs to operate at Rb transition energy or laser frequency.
This is the 780 nm range.
The laser should achieve a mode-hop free region of all 8 GHz. This is a secondary requirement to attempt a sweep large enough to cover Rb absorption scans with the Littman configuration. At the very least, the laser should be able to tune its frequency without mode-hopping within a range of frequency to cover one of the re-pumps of Rb.
The enclosed GSDL should be as compact as possible to minimize space on optics lab bench.
No component should have to be changed once all components
(mirror, grating, and diode clamp) have been bolted down.
The GSDL needs to be completely enclosed to prevent temperature change from air flow.
2.3.2
Initial Design Proposal
19
2 | D ISCUSSION
– Design Proposal
2.3.3
Pre-Construction Changes to Design
The original schematics were given to Bruce Kapplauf to review before the GSDL construction. As instructed by Bruce Kapplauf, a few changes had to be made. These changes included:
1.
Laser is preferred to beam out normal to the side plates such that it is easier to position and clamp down on the optics table when in use. Therefore the box orientation had to be rotated 40º counterclockwise.
2.
Make grating mount thicker from 0.5 in to 0.75 in. This is to use larger sized screws; an 8-32 threaded screw because the QDG lab had more supply of this screw.
3.
The bottom of the mirror and grating mount had an indented relief (see Appendix).
These reliefs were changed from a rectangular stand to only two sided reliefs to minimize area of contact. This is to prevent uneven surface on the base-plate changing the angle of the mirror or grating.
4.
The mirror mount screws were changed from 6-32 threaded to 8-32 threaded and the depth of the holes were changed from ¼ in to ½ in to improve stability.
The final schematics are attached in Appendix. The resulting design is shown in Figure
(above).
20
2 | D ISCUSSION
– Data
2.4
Data
The following is a set of recorded data on optimized frequency settings of the laser for different occasions.
Date mm-dd-yy
01-22-07
01-23-07
01-23-07
01-26-07
02-01-07
Event
Optimizing with amplifier on the
PTZ driver
Optimizing with amplifier enclosed for PTZ driver
Aligning height of mirror and grating
Minimize modehopping for lower temperature and current
Initial attempt to optimize Rubidium cell curve
Current
73.8 mA
80.8 mA
69.9 mA
47 mA
62 mA
78 mA 02-01-07 Resetting
02-20-07
02-27-07
02-27-07
Adjusted optical component settings
Optimization based on 1 st order beam position
Optimization based on 1 st order beam position
02-28-07 Low Grating Angle
77.2 mA
80.6 mA
76.4 mA
92.7 mA
02-28-07
03-02-07
03-02-07
03-26-07
Feed Forward on
Laser Current
Optimization with
Feed Forward
38º Diffraction angle
Lost Optimization, tried again
03-26-07 Attempt 2
92.7 mA
75 mA
90.9 mA
79.6 mA
81.2 mA
T set
19.175 kΩ
18.149 kΩ
11.776 kΩ
18.387 kΩ
18.163 kΩ
18.168 kΩ
18.243 kΩ
18.531 kΩ
18.237 kΩ
18.357 kΩ
18.357 kΩ
17.688 kΩ
18.083 kΩ
18.06 kΩ
18.07 kΩ
T act
19.170 kΩ
18.149 kΩ
11.776 kΩ
18.387 kΩ
18.163 kΩ
18.168 kΩ
18.239 kΩ
18.536 kΩ
Bandwidth
Not Taken
Not Taken
Not Taken
Not Taken
Approx. 3
GHz
Not Taken
Approx. 2
GHz
Approx. 2
GHz
18.237 kΩ
18.355 kΩ
18.357 kΩ
17.684 kΩ
18.082 kΩ
18.06 kΩ
18.07 kΩ
Approx. 2
GHz
Approx 4
GHz
Approx 10
GHz
Approx 6
GHz
Approx 7
GHz
Approx 4
GHz
Approx 4.5
GHz
21
2 | D ISCUSSION
– Results
2.5
Results
The project yielded many results in the different phases. Each step in optimization yielded different graphical data that had to be interpreted in order to understand the method required to further maximize frequency bandwidth of the laser.
2.5.1
Discussion of Project Data
Florescence of Rubidium Cell
The Ramped PTZ sweep of Rubidium Absorption Spectrum
22
2 | D ISCUSSION
– Results
23
2 | D ISCUSSION
– Results
Bandwidth Optimization with Beam Position
24
2 | D ISCUSSION
– Results
2.5.2
Discussion of Project Issues
Need filler text here.
25
2 | D ISCUSSION
– Final Design
2.6
Final Design
Need filler text here.
2.6.1
Finalized Laser
The complete Littman laser is enclosed into a rectangular box with dimensions of # x # x #
(inches). It is composed of an aluminum base with plexiglass side and top plates. Three components are mounted onto the base of the slave laser: grating mount, mirror mount, and a laser diode clamp as shown in the figure below.
The input ports are: laser diode current control, laser temperature control, and a piezoelectric driver source.
26
2 | D ISCUSSION
– Final Design
2.6.2
Post-Construction Changes to Design
A few changes were made in order to optimize the laser frequency bandwidth after all components were bolted down onto the base plates. These changes are listed as follows:
The base-plate has been rotated such that the short sides of the sideplates are the input (temperature control, current control, piezoelectric driver) and outputs (laser beam).
Mirror mount uses only one threaded hole instead of both holes (as can be seen from the schematics in the appendix.
Mirror mount stand is rotated backwards such that the complete mirror mount on top is pushed back and away from the diffraction grating to increase the cavity length.
The side plate behind the Mirror mount had to mill an indent to allow spacing for the Knobs on the Mirror Adjustment mount. A mistake was made during machining so a hole was made to clear out the indent and a glass slide is glued on to seal off the box.
27
2 | D ISCUSSION
– Laser Locking Results
2.7
Laser Locking Results
Pavle, can you do this part here?
28
3 | C ONCLUSION
3.0
C ONCLUSION
Pavle, can you could just generalize and give a summary of the laser locking here. I’ll fill in the rest of the stuff to make the conclusion more complete.
29
4 | R ECOMMENDATIONS
4.0
R ECOMMENDATIONS
Pavle, I don’t think this part has to be long, maybe just a short pargraph I believe.
30
5 | A PPENDICES
5.0
A PPENDICES
31
6 | R EFERENCES
6.0
R EFERENCES
[1] Park, S.E, T. Kwon, E. Shin, H. Lee. 2003. A compact extended-cavity diode laser with a littman configuration. IEEE 0018-9456/03.
[2] Gawlik, W. J. Zachorowski. Stabilization of diode laser frequency to atomic transitions. Optical Applicata Vol XXXIV, No. 4, 2004.
[3] Kraft S., A. Deninger, C. Trück, J. Fortágh, F. Lison, C. Zimmermann. Rubidium spectroscopy at 778–780 nm with a distributed feedback laser diode. Wiley
InterScience. Laser Phys. Lett. 1–6 (2004) / DOI 10.1002/lapl.200410155.
[4] Pedrotti, F., L. Pedrotti, and L.M. Pedrotti.
Introduction to Optics 3 rd Edition.
Prentice Hall. 2007: 229-230, 614-616.
[5] Truscott, A.G., N. Heckenberg, H. Rubinsztein-Dunlop. 1999. Frequency stabilised grating feedback laser diode for atom cooling applications. Optical and Quantum
Electronics. 31 : 417-430, 1999.
[6] Wieman, C., L. Holberg. 1991. Using diode lasers for atomic physics. Rev. Sci.
Instrum.. 62 (1)
[7] Wikipedia: Laser Cooling. Retrieved 1 st November 2006 from the World Wide
Web: http://en.wikipedia.org/wiki/Laser_cooling
[8] FU Berlin: Chapter 7 - Pump probe spectroscopy of Ultra-cold Rb
2
Molecules.
Retrieved March 29 th , 2007 from the World Wide Web: http://www.diss.fuberlin.de/2007/ 104/chapter7.pdf
32
P ROJECT E VALUATION
P ROJECT E VALUATION
This project went very bad.
33
