Experimenters involved in measuring single-shot, fast-transient phenomena are fast turning toward photonics to meet their experimental needs. Photonics is the technology of generating and using light in detection, diagnostics, communications, and information processing systems. This technology includes light generation, transmission, deflection, amplification, detection, and recording and includes areas of lasers and other light sources, fiber optics, electro-optical instrumentation, and optical components and instrumentation. Photonics systems offer immense advantages and possibilities over ,Systems that use electrical signals. The advantages realized by photonics systems include increased data quality (in terms of bandwidth, resolution, reliability, and often accuracy and precision), increased data quantity, and economic and operational benefits. Single-shot, fast--transient analog measurement techniques introduce some special considerations and challenges for the experimenter that are not encountered in telecommunication applications of photonics. This overview explores some of these issues with emphasis placed on photonic sensors and data recorders. A sampling of photonics diagnostic systems is presented along with a discussion of future possibilities and challenges facing the experimenter. This exciting new field is in its infancy, and many of the diagnostic techniques presented are still evolving into yet more powerful tools for the experimenter. Single-shot, fast-transient analog photonics diagnostic techniques have experienced an almost exponential growth over the past few years and are expected to have a major impact on governmental, industrial, and academic communities involved with such measurements.
We have successfully recorded analog signals, using the High-Speed Multi-Channel Data Recorder1 (HSMCDR), where the signal source and the recorder were required to be -2000 m apart. In addition, it was necessary for the system to operate remotely. This photonic data recording experiment consisted of a HSMCDR, laser diode transmitters and the -2000m fiber link between them. Of six transmitters, one was used to provide the HSMCDR trigger. Another transmitter sent the trigger waveform to be recorded by the HSMCDR. The remaining four transmitters converted analog electrical data signals into analog photonic signals for transmission to the HSMCDR. Since all signals were coincident, the analog signals were sent over an additional 200 m of fiber to obtain a 1 μs delay, necessary to allow for delay from trigger pulse to start of streak. The system bandwidth was about 350 MHz. The HSMCDR performed all its trigger monitoring, shutter control, arming, and data transfer functions under the direction of a computer. To assess system performance, we designed and installed self-diagnostic circuits that recorded key timing signals on a fast strip-chart recorder. Although we were only provided with five data channels, the HSMCDR used is a 20-channel system. We assembled this system in two days and were ready for data recording within one week. We will describe the system and present our results. In addition, a second generation HSMCDR built around a streak camera with a large area photocathode and charge-coupled device (CCD) digital readout will be described. This system will record up to 40 channels of data.
The frequency response of data links transmitting analog signals over kilometer-length graded-index optical fibers is limited to 1 GHz by fiber bandwidth. System bandwidths can be substantially increased if single-mode fibers are used. In the developmental system presented here, a 1-km single-mode optical fiber is used to transmit an intensity-modulated 810-nm optical signal to a streak camera recorder. Useful system bandwidth could extend to 5 GHz.
Scintillator, fiber, and detector technologies have progressed over the last few years so that it is now possible to record radiation images over optical fiber bundles 1 km in length. Scintillators that emit at 730 nm will be coupled with wavelength-multiplexing techniques to telecommunication fibers for transmission to a demultiplexer and detection by GaAs microchannel-plate image-intensifier tubes. The technique and design considerations, such as time-integrated versus time-resolved images, resolution and system dynamic range will be described.
A photonic data recorder (PDR) is currently being developed that uses light rather than electrical signals to record physical measurements. The device operates much like a conventional streak camera but with the streak tube replaced by an electro-optic crystal. The crystal deflects the light directly, eliminating the need for a photocathode. The usable wavelength range of the PDR extends well beyond the 0.5 - 0.8 cm range of the S20R photocathode typically used in streak tubes. The spectral region is determined by the transmission range of the electro-optic crystal and the responsivity of the photodiode detector. For example, lithium niobste (LiNb03) transmits between 0.35 and 0.2 cm, and silicon photodiodes respond from 0.3 to 1.1 cm. One of our primary signal sources is a commercial single mode laser diode operating at - .82 that falls within this response range. Our goal is to develop a single channel instrument with: o S/N > 30-for optical signals from .4 to 1.1 μm. o resolution of 50 - 100 (or better) resolvable spots per sweep. o substantial cost reduction (2 - 5x) over presently available digitizing recorders. This instrument incorporates most of the features found in the conventional digital and analog oscilloscope recorders.
Spectral-streak equalization is a technique that has been developed to compensate for the material dispersion in optical fibers when used in conjunction with an electronic streak camera. Material dispersion occurs because different wavelengths of light travel at different speeds through glass fibers; the resulting difference in transit time broadens light pulses, and can lead to errors in high bandwidth photonic measurements. An instrument designed to compensate for this effect has been in use for the past several years in systems used to evaluate underground nuclear tests. A new instrument has been developed that has the following advantages: it can equalize several channels with one set of optics; it uses considerably less space; it has better resolution and greater efficiency, and it is more cost effective. This paper reviews the basic principles, describes the equalizers currently in use, discusses the design considerations of the new equalizer, describes a prototype four channel instrument, details efficiency estimates, outlines calibration procedures, and gives test results.
Since the introduction of the High-Speed Multi-Channel Data Recorder (HSMCDR)1 various advances have been made in improving its signal reproduction quality while maintaining its high bandwidth and channel density. The second generation HSMCDR II system has recently shown a 3.5 GHz, 3 dB bandwidth single-shot recording capability in 25 simultaneous channels. Due to improved analog laser diode technology and solid state image readout, the recorded signal accuracy and reproducibility of the HSMCDR has improved substantially. A detailed system characterization has been performed with emphasis placed on the practical issues of operating the HSMCDR system. Proper operation of the recording system requires the operator to fully understand the variety of constraints inherent in several of the system's components. Operation within these constraints has shown the HSMCDR II to have an unequalled price-performance factor.
Due to the unusual experimental conditions experienced at the Nevada Test Site (NTS), unique requirements are imposed on calibration systems. This paper describes a fiber optic system which was used to calibrate a recent experiment. Eight photodiodes (PD's) were used as signal detectors. The experiment required that the relative timing between the signal channels be known to +50 ps. A multiple fiber system was used to provide sufficient signal amplitude to the detectors without stimulating Raman. By the use of this calibration tech-nique, relative timing to +35 ps was achieved.
A technique for in-fiber amplification of single transient analog optical signals has been investigated. It achieves high gain and linearity in single mode and multimode fibers using the backward stimulated Raman process.
Many photonic sensors produce signals that are not a linear function of the stimulus. For example, radiation sensors' may produce an exponential decay response. Similarly, the Faraday current2, microwave3, Kerr voltage4, and VISAR5 sensors all produce a sinusoidal response. A perfect recorder system would reproduce the signal exactly. However, all real recorders introduce noise and bandwidth attenuation. The bandwidth response, dynamic range, and signal-to-noise ratio of the decoded data depend on the manner in which the data are encoded. This paper will analyze the various encoding techniques and their effects on data recording. Both simulated and experimental data will be presented which illustrate the effects of encoding on the dynamic range, signal-to-noise ratio, and bandwidth of a signal. The data will illustrate the advantage of recording phase encoded signals.
A fiber optic electric field sensor for lightning research measurements based on the transverse electro-optic effect in Bi4(Ge04)3 electro-optic crystals (BGO) is described. The sensor is completely dielectric in composition and can monitor both electrib field strength and direction. The described sensor has a linear sensing range between 100 V/m and 1 x 107 V/m and an AC measurement bandwidth capability which approaches a 1 GHz which is more than adequate for lightning research. The effect of electro-optic crystal geometry on fiber optic sensor measurement range and measurement bandwidth is discussed. A method of utilizing a single electro-optic crystal to make two axis electric field directional measurements is described. A fiber optic electric field sensor configuration which incorporates two electro-optic crystals, and which is capable of measuring electric field direction is described. Sensor test data generated with both DC and AC applied electric fields is presented.
This paper presents computations of the electric polarizability of a uniform isotropic dielectric parallelepiped, obtained by an application of the Moment Method, over a range of relative permittivities from 5 to 100, and relative dimensions from lxlxl to lxlxl4 and from lx2xl to lx2x14, where the electric field is applied along the major axis. The results are useful in quantifying the sensitivity of and far-field scattering by an electrically small electric-field probe employing direct field coupling to an electro-optic crystal. For example, a crystal with dimensions 1 mmxl mmx10 mm and a relative permittivity of 50 has a computed polarizability of 195.mm3 along the major axis, which corresponds to an average internal electric field intensity of .398 times the applied field intensity.
A photonic electric field probe using the Pockels effect in bulk LiNbO3 is used to measure electromagnetic fields from 10 to 100 V/m. The observed frequency response is flat up to 1.6 GHz and extends beyond 2 GHz. Over the majority of the frequency range, field strengths down to about 6 V/m would be detectable above the noise floor when using a 3 kHz detection bandwidth. Present experimental results indicate a linear dynamic range of 30 dB for the probe. Increasing the optical carrier power and lowering the noise floor are expected to improve the dynamic range to above 50 dB.
There is increasing interest in LiNbO3:Ti integrated optical devices as transducers in fast, analog, data-acquisition systems. A possible limitation to such use, however, results from the photorefractive effect in LiNbO3:Ti. From existing data, we know of difficulties due to the photorefractive effect with continuous-wave (CW) illumination. Here we present pulse measurements at 810 nm of the optical-transmission distortion due to the photorefractive effect in LiNbO3:Ti waveguides. We subjected the optical devices to pulsed input light generated by a mechanically shuttered dye-laser system. We then measured and compared the transmitted light to an input monitor and obtained the transmission of the LiNbO3:Ti devices as a function of time. We varied the peak power from 50 pW to 150 mW, and the pulse length from 10 ms to 1 s. We have obtained a 21-mW output with a 10-ms duration from a LiNbO3:Ti waveguide with no prom2t transmission degradation. Further, we have found preliminary evidence suggesting an activation time of 100 to 150 min. for the onset of photorefractive transmission degradation.
Directional coupler modulators (DCMs) have gained interest as a method of obtaining high-speed analog modulation of large signals and have the potential to overcome difficulties encountered with schemes that modulate the light source directly. Recording precise, high-bandwidth analog information above approximately 3 to 4 GHz requires a streak camera. Photocathode sensitivities are extremely low at 1300 nm, where the bulk of the DCM work has been done to date. We are, therefore, developing a DCM to operate at 810 nm and to approach bandwidths of approximately 15 GHz. We present measurements of a systematic parameter study of the optical coupling properties, the electro-optical transfer function, and the single-shot, pulsed, electro-optical response of these devices. We discuss applications to some high-speed diagnostic systems.