LASER SHUTTERS: Shutter technology keeps pace with laser advances
DAVID C. WOODRUFF
High-irradiance lasers can be found in a vast range of applications from semiconductor fabrication to the laser-guide-star adaptive-optics system at the Keck Observatory (Mauna Kea, HI). These lasers are extremely versatile tools, but they are potentially dangerous. In all cases a highly reliable means of beam termination must be available. In the case of semiconductor fabrication, the beam must be shut down immediately if there is a safety breach of the system. In the case of the observatory, the beam must be shut down, for instance, if an aircraft approaches. The shutdown function is generally performed using a laser shutter.
When the shutter is open, the beam travels through undisturbed. Closure completely blocks the beam (see Fig. 1). During closure, the energy of the beam is diverted into an integral light-absorbing baffle that can, in some units, heat-sink beams in the kilowatt range indefinitely. There is thus no need to power down the laser. The beam is modified only during the switching transitions that typically last hundreds of microseconds.
Laser-shutter applications are not limited to switching high-irradiance laser beams, they are also used to pass low-level light and even to block flying debris. For example, in lidar systems, shutters are used to block the receivers to protect photosensors from being overdriven by the initial backscatter of the outgoing laser pulse. After the initial pulse, the shutter quickly opens to catch the low-level returning light. Such shutters are used when researching x-ray spectra from pulsed laser targets. The shutter closes fast enough to prevent debris from the exploding target from reaching the x-ray spectrometer. Modern shutter technology can deliver bursts of laser energy at rates as fast as 500 Hz, with periods as short as a millisecond. The duration of such bursts is also highly reproducible.
Beyond solenoids
Early laser shutters used solenoids. Typically, a rotary solenoid and spring were provided to rotate an aperture in a metal plate through the laser beam. Lubricated bearings were provided to keep the friction low. To minimize vibration and the size of the solenoid, the inertia had to be kept low. This meant that the metal plate had to be thin, which lessened the plate's ability to sink heat. Heat was thus conducted to the bearings and increased the outgassing of wet lubricants, which became gummed with foreign particles. Dry lubricants could not be used because they contributed to particle debris. Even with low-inertia systems, vibration introduced by the acceleration and deceleration of the solenoid at the end of its stroke was considerable. At best, such shutters had life spans on the order of 100,000 to a million cycles.
Modern laser shutters have no bearings and require no lubricant. The only moving part is a low-mass, flexure mirror assembly consisting of a flexible, ferromagnetic cantilever membrane that is moved in and out of the beam by an electromagnet. When this assembly is in the laser-beam path, a mirror in the flexure diverts almost all of the laser energy into an integral light-baffle heat sink (see Fig. 2).Geometry, surface morphology, and the atomic material of the light baffle guarantees that most of the radiant energy is absorbed and passed to the heat sink. Conduction cooling is achieved through mechanical mounting to a large mass. Some high-energy lasers require a water-cooled heat sink, such as a water-circulating "chiller plate" that can easily be attached to the heat-sink mass.
To make the shutter fail-safe (shut down when power fails) the highly reliable cantilever-flexure closes when power is removed from the electromagnet. In some units, switching is accomplished in less than 200 µs and any bounce of the flexure is designed well out of the beam area, ensuring complete on-and-off operation. The fail-safe flexure mechanism can be interlocked with the rest of the system, such that any faulty device creates an open loop, ultimately dropping the voltage to the controller input.
Modern shutter designs are also scalable. Shutters with larger apertures have longer switching periods, which vary with mass. Attachment of dielectric optics can increase the power handling but slow the acceleration. Typically, the flexure passes through the beam at about 2 mm/millisecond, but it can reach speeds of 6 mm/ms (6 m/s).
An aluminum-coated mirror on the flexure surface is often sufficient to divert the energy of a wide range of laser beams with various wavelengths. Gold mirrors are an option for infrared beams. Dielectric mirrors on attached glass substrates are provided for high-energy lasers at specific wavelengths. The mirrors are designed to maximize reflectivity when the flexure is closed, providing a high damage threshold. For metal mirrors, an arrow marked on the shutter's input aperture indicates the direction of the beam's electric vector, facilitating proper alignment with the laser. The flexures are designed to move rapidly in response to the switching force of the magnetic field, to conform in shape and provide an intimate fit with the surfaces of the electromagnet poles at the beam exit, and to be rigid enough to present an essentially flat surface to the beam.
Electromagnets are wet-wound in epoxy resins that ensure outgassing levels low enough to surpass NASA outgassing standards. Gapped toroidal shapes and closed-path couplings minimize flux leakage. A thin sheet of steel is usually more than adequate to isolate even the most magnetically sensitive devices such as faraday rotators.
The catenary shape of the electromagnetic poles assures equal force on the flexure mirror assembly when accelerating. Since forces are distributed equally across the flexure, contact with the pole is nearly simultaneous everywhere, with minimum pressure points. Equal distribution of the force also ensures intimate contact of the flexure with a pole at the end of travel.
Heat management of the magnet is accomplished by using thermal-conducting epoxy and a metal path to the base plate of the shutter (see Fig. 3). After the flexure reaches the pole, the controller delivers only a small holding current. Keeping the magnet cool ensures that the resistance of the coil doesn't change, enabling the current and switching times to remain constant. Heat management of the magnet also greatly reduces thermal air gradients introduced into the beam's path.Controllers drive the magnet. A simple capacitor discharge controller is adequate for simple shutters, and offers advantages as a result of having few parts, including low cost, high reliability, and a very long mean time between failures. The capacitor discharge controller can easily be constructed by the OEM as an integral part of the product.
More-sophisticated controllers deliver a sculptured, voltage-regulated waveform to drive faster shutters. This waveform is designed to rapidly accelerate the flexure and bring it to rest with sufficient holding current. Slight overdamping of such systems eliminates ring and reduces bounce for high-speed closure. Close proximity to critical damping also minimizes energy transferred to the magnet, further enhances heat management, and greatly reduces mechanical shock and vibration.
What is the future of electromechanical laser shutters? As lasers get bigger and better they can only place greater demand on shutters. Engineers at nmLaser are developing a shutter with a noncontact flexure designed to close at a position slightly away from the pole, essentially floating in air to eliminate any potential vibration or bounce caused by rapid deceleration of even the low-mass flexure. Slightly overdamped controllers will eliminate system ring, ensuring efficient energy transfer to the magnet. Such flexures will be unaffected by foreign particles clinging to the magnetic pole. They will also be suitable for use in clean rooms and offer much longer lifetimes than is currently available.
David C. Woodruff is the president of nmLaser Products, 337 Piercy Road, San Jose, CA 95138; e-mail: [email protected]; www.nmlaser.com.