Transcript Slide 1

Stabilizing a He-Ne Laser by Thermal Control of Cavity Length
Ewuin Guatemala, Samuel Goldwasser*, Martin Cohen, and John Noé
Laser Teaching Center, Department of Physics & Astronomy
Abstract
Measurements
Precision optical experiments require lasers whose frequency and/or intensity remain
highly stable over time. Common sealed tube helium neon (He-Ne) lasers do not fulfill
these conditions unless they are modified. He-Ne laser cavities between 20-30cm in
length produce two or three longitudinal modes that are amplified by the Neon gas'
Doppler-broadened gain profile (1.5 GHz in width). The spacing between adjacent
modes is related to the cavity's length (L) by the equation ν = nc/2L where ν is
frequency, c is the speed of light and n is an integer corresponding to the number of
modes oscillating in the cavity. However, as the laser heats, the length of the tube
expands resulting in mode sweeping, that is, the position of the modes shifts relative
to the gain profile causing fluctuations in output intensity and frequency. One
approach to stabilize a He-Ne laser is to keep the thermal expansion of the tube to a
minimum using a thermal feedback circuit.
The intensity of the laser is tracked from startup to finish with a
photodetector which produces a current when the beam is upon its
active area. The current is converted to a voltage by placing a resistor
between its two leads.
Before locking the laser, the setup is set to “pre-heat” mode for at least
five minutes before feedback is introduced into the system. At the end of
preheat (around the 600th second), the circuit is switched to “lock” mode,
and stability is usually reached within a few seconds. The laser intensity
is measured from the beginning of the operation and is tracked for about
30 minutes into lock mode. The intensity of the laser is measured both
with and without a polarizer in front of the photodetector.
The error signal and heater voltages are taken directly from the PCB.
All voltages are interfaced to a computer to track these values over
time.
How Laser Intensity and Error Signal Relate
Intensity without a polarizer
7
3.6
Detector Voltage (V)
6
Voltage (V)
Since adjacent modes produced by the laser are orthogonal, they can be segregated
with a polarizing beam splitter (PBS), and their relative intensities serve as indicators
of how the modes are sweeping underneath the gain profile. A proportional-integral
(PI) circuit can be used to regulate the relative intensities of the two modes, locking
them in place. This setup was successfully completed and tested, achieving
frequency stability to an estimated 25MHz, or 55ppb of the He-Ne laser’s 474THz
(632nm wavelength) output, and cavity length stability of 12nm.
Stabilization Results
5
4
3
2
1
Caption: The gain profile of Neon gas is Gaussian in
shape and is 1.5 GHz in width. The longitudinal
modes are extremely narrow in comparison, and are
separated by 645MHz for a 24cm length cavity. The
intensity with which each mode is amplified by the
medium depends on its position underneath the gain
profile (it is highest at the center and smaller at the
ends).
0
10
20
30
40
Time (s)
50
60
Photodetector
70
3.3
3.2
3.1
3
2.9
80
0
500
1000
1500
2000
2500
Time (s)
Error Signal
This graph is taken with a polarizer placed in front of the photodetector to
isolate one mode, producing the largest contrast between minimum and
maximum intensity. Each extreme is an indication that one mode is
predominantly lasing, and should correspond to an extreme in the error
signal. As seen in the graph, this is the case, demonstrating the proper
integration of the optical (beam splitter and photodiodes) and electronic
components of the setup.
The intensity of the laser slowly builds after startup and is clearly oscillating
over a wide range in comparison to its stabilized intensity. Once in lock
mode, the rms fractional voltage was cut to 1%.
Intensity with a polarizer
Detector Voltage (V)
Setup
Behavior of the Setup During Lock
10
9
Voltage (V)
3.4
2.8
0
The He-Ne laser stabilized for the project was a Spectra-Physics model 088, 24cm
in length with 1mW of output power. The heating coil was made from 190 feet of 30
AWG (20-ohm total resistance) enamel-insulated magnet wire obtained from a chain
electronics store. The wire is wound bifilar around the tube to avoid inducing
magnetic effects, and adheres to the tube with epoxy at a few locations.
3.5
8
7
2
1.8
1.6
1.4
1.2
1
0.8
0.6
0.4
0.2
0
0
6
500
1000
1500
2000
2500
Time (s)
Bare tube.
5
4
0
Tube with heater
with the back
beam aligned with
the PBS
Photo of the PBS with the pair of
silicon photodiodes. Two
orthogonally polarized beams
emerge from the PBS at right
angles. A mirror reflects one of the
beams towards the photodiode.
The laser’s back beam is coupled to a polarizing beam splitter (PBS) which
produces two beams with orthogonal polarizations. Each beam goes into a
photodiode, producing two currents which are compared by the proportional
part of the PI circuit producing an amplified error signal. The integrating
portion of the circuit sums the error signal over time and has the effect of
adding stability to the feedback circuit (a feedback circuit with only a
proportional term provides a signal that oscillates before it settles; the
integral term has a damping effect on this oscillation).
10
Time (s)
20
30
40
Detector Voltage X 8
50
60
Error X 1.6
70
80
Heater
The apparatus is constantly adjusting to changes in temperature and cavity
length throughout the time it is locked. The intensity of the laser is still
varying minutely when in lock mode, although these fluctuations are small in
comparison to the intensity shifts when unlocked. The setup is sensitive
enough to produce an error signal closely following these changes in
intensity. For example, as the intensity fell in second 40, the error signal
rose, prompting an increase in current across the heater.
The inclusion of the polarizer maximizes the difference between intensity
minima and maxima, which in this case resulted in a rms fractional
voltage of 52%. In lock mode, this value diminished to 4%. This value can
be used to estimate the frequency and cavity length stability achieved by
the setup.
Stability Estimates
Making estimates of the frequency and cavity length stabilities is made
relatively simple by using the data from the polarized lock. A change from a
minimum to a maximum voltage (or vice versa) represents a switch from one
mode to another, which is equivalent to a free spectral range (c / 2L) which in
this case is 645Mhz. If the voltage is fluctuating only 4%, this means that the
frequency is fluctuating 25Mhz, which in comparison to the frequency of the
He-Ne Laser (474Thz) is 55ppb.
Each mode switch also corresponds to an increase or decrease in the cavity
length of a half wavelength of the light (316nm). Once again, a 4% variation in
this value is 12nm, meaning that the length of the cavity can be controlled on a
very small scale for slow fluctuations in cavity length.
Left is a Scanning Fabry-Perot Interferometer (SFPI) whose cavity length is varied by a PZT
driven by a function generator. It also contains a photodiode to measure the intensities of
the light resonating within the cavity. The voltage across the photodiode can thus be tracked
on an oscilloscope (right) to display the longitudinal modes within the cavities and their
relative intensities. Each pair of peaks in the photo on the right represents one of the modes
of the laser light resonant in the SFPI. These peaks rise and fall indicating the “sweeping” of
the modes across the laser’s gain profile; they hold steady when the laser is locked.
This project demonstrates how a simple thermal feedback method can be
used to reduce the intensity and frequency fluctuations of a sealed tube He-Ne
laser within minutes, and can be controlled over long periods of time. Further
work can be done to further limit frequency and cavity length fluctuations such
as the use of insulation. Or, some of the principals learned here can be
applied to the stabilization of other laser systems.
References
A. Lindberg, “Mode frequency pulling in He–Ne lasers
American Journal of Physics,” Am. J. Physics. 67, 350 (1999)
The signal from the integrating op-amp
goes to a transistor which powers the
heater. The amount of current going to
the heater is constantly being adjusted
according to the difference in intensities
of each mode. The trim-pots associated
with each op-amp can be adjusted to
achieve the desired ratio of mode
intensities.
Control Tutorials for Matlab and Simulink: PID Tutorial , April 2010
http://www.engin.umich.edu/class/ctms/pid/pid.htm
F. Stanek, R.G. Tobin, C.L. Foiles, “Stabilization of a multimode He-Ne laser: A vivid demonstration of
thermal feedback,” Am. J. Physics. 61, 932 (1993)
G.A. Woosley, M.Y. Sulaiman, M. Mokhsin, “Correlation of changes in laser tube temperature, cavity
length, and beam polarization for an internal-mirror helium-neon laser,” Am. J. Physics. 50, 936 (1982)
S. Goldwasser, Sam’s Laser FAQ, April 2010. http://www.repairfaq.org/sam/lasersam.htm
The PI circuit is contained on a small 4cm X
5cm printed circuit board. The wires in a
spare Ethernet cord are used to connect the
PCB to the photodiodes and heater.
A picture of the setup shows the laser wound
with the red heater coil, and the yellow
Ethernet cord connecting the photodiodes to
the circuit in the background.
* Dr. Goldwasser, of Bala-Cynwyd, PA, is the author of the widely-known
internet resource Sam's Laser FAQ. He provided the laser tube and circuit
board for this project, and the Fabry-Perot analyzer. Further information
about the circuit design may be found at the Laser FAQ web site.