Abstract

Stabilizing a HeNe laser by thermal control of its cavity length

Ewuin Guatemala, Samuel Goldwasser*, Martin Cohen, and John Noe

Laser Teaching Center, Department of Physics & Astronomy

Precision optical experiments require lasers whose frequency and/or intensity remain highly stable over time. Such stability can be achieved in the familiar sealed-tube helium-neon laser through the use of a feedback circuit that controls a heater applied to the glass laser tube. Depending on its tube (cavity) length a HeNe laser produces two or more closely spaced frequencies (modes) over a range of about 1.5 GHz set by the Doppler-broadened gain medium. While this range is just 3 parts per million of the average output frequency of 474 THz (wavelength lambda = 632.8 nm), the frequency spread of each individual mode is roughly a thousand times smaller, or a few parts per billion. As the laser cavity length changes with changing temperature, the center frequencies of these "sharp" modes shift across the gain profile, causing dramatic mode intensity variations as well as frequency shifts. The fact that adjacent modes can be distinquished by their orthogonal polarizations allows a suitable circuit to regulate the relative intensity of two adjacent modes by thermal feedback. The feedback effectively locks the modes to fixed positions on the neon gain curve, thereby stabilizing both their frequency and intensity.

The HeNe laser we stabilized (Spectra-Physics model 088) is about 23 cm in length, and has a mode spacing of 641 MHz. The weak (< 10 microWatt) "back" beam of the laser was passed through a polarization-sensitive beam splitter cube to separate the orthogonal modes and direct these to two small Si photodiodes. The proportional-integral (PI) feedback circuit is contained on a small (4x5 cm) printed-circuit board. It includes a preheat-off-lock switch, two LEDs to indicate the relative light intensity on the two photodiodes, and a third LED to indicate relative heater power. The circuit is powered by a single 12 Volt DC volt power supply. The laser tube heater was fabricated by winding 190 feet of 30 gauge enamel-insulated magnet wire obtained from Radio Shack around the tube. The total resistance was 20 ohms, close to the optimum value, and the wire diameter was such that the entire tube surface was covered, thereby optimizing the speed of the thermal response. The wire was wound as two parallel strands starting from the overall midpoint; this bifilar winding technique eliminates magnetic fields that would otherwise be produced in a coil. During the winding process the wire pairs were anchored to the tube in a few places with epoxy. The wound tube was left open to the air to enhance heat dissipation.

Locking the laser is straight-forward. The laser is turned on, and the circuit set to preheat mode. After about 10 minutes the circuit is switched to lock mode and lock routinely follows within a few seconds. The stability of the locked laser was qualitatively confirmed by observing the two modes with a Fabry-Perot analyzer. Quantitative estimates of stability were made by recording the intensity of the full laser beam with a photodetector. Readings were taken once per second for periods up to an hour using a USB-1208LS data-acquisition device from Measurement Computing Corp. Measurements were made both with and without a polarizer in front of the detector. With the polarizer in place and suitably oriented only one mode contributes and the intensity fluctuations are much more extreme than when the summed intensity of both modes is recorded. The rms fluctuations over 30 minute periods were measured to be 4% and 0.9% with and without the polarizer, respectively. The 30-minute frequency stability can be estimated from these results to be better than 25 MHz, or at least 100 times better than that of the unlocked laser. Similar recordings were made of the heater voltage and the photodiode error signal as a function of time. The heater voltage during lock is typically about 60% of the preheat value.

* Dr. Goldwasser, of Bala Cynwood, 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 in the Laser FAQ.