SEARCH · Engineering Papers
Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.
Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.
Circuit protects all output lines of switching regulator against overloads without requiring current sensors on every line. If overload is sensed, device short circuits bias on switching transistor so that power is rapidly cut off from loads. Circuit also includes delay network to inhibit erroneous operation during startup.
Highly-reliable lightweight solid-state RF (radio frequency) transfer switch replaces less reliable mechanical switch in handling medium power for S-band power amplifier.
Packaging approach makes resistors more accessible and easily identified with specific switches. Failures are repaired more quickly because of improved accessibility. Typical board includes one resistor that acts as circuit breaker, and others are positioned so that their values can be easily measured when switch is operated. Approach saves weight by using less wire and saves valuable panel space.
A method for modelling switching converters in the discontinuous conduction mode is developed, whose starting point is the unified state-space representation, and whose end result is a complete linear circuit model which correctly represents all essential features, namely, the input, output, and transfer properties (static dc as well as dynamic ac small signal). While the method is generally applicable to any switching converter operating in the discontinuous conduction mode, it is extensively illustrated for the three common power stages (buck, boost, and buck-boost). The results for these converters are then easily tabulated owing to the fixed equivalent circuit topology of their canonical circuit model.
A novel switching dc-to-dc converter is presented, which has the same general conversion property (increase or decrease of the input dc voltage) as does the conventional buck-boost converter, and which offers through its new optimum topology higher efficiency, lower output voltage ripple, reduced EMI, smaller size and weight, and excellent dynamic response. One of its most significant advantages is that both input and output current are not pulsating but are continuous (essentially dc with small superimposed switching current ripple), thus resulting in a close approximation to the ideal physically nonrealizable dc-to-dc transformer. The converter retains the simplest possible structure with the minimum number of components which, when interconnected in its optimum topology, yield the maximum performance.
The design of a packet switch incorporating a minicomputer to control and process data packets in a communications system has been reported. This paper proposes a design which uses a 6500 series microprocessor to realize the functions of a message switch in order to evaluate the viability of microprocessors in this application. The proposed design realizes the functions of header analysis, packet security, error control and line prioritization. The system hardware and software are presented together with a brief description of operation.
A small signal average model is derived to study analytically the complex interaction among input filter, output filter and control loop, which frequently causes degradation of switch regulator performance. Analytically based design guidelines are formulated, and key design parameters are identified. It is shown that minimization of the forward transfer characteristics and the output impedance of the input filter at filter resonance are key to designing an input filter for a switching regulator with given output filter parameters and specified line and load conditions.
A coaxial switch capable of operating in a vacuum with high RF power in the 1.2 GHz range without multipactor breakdown, and without relying on pressurization with an inert gas is described. The RF carrying conductors of the switch are surrounded with a high grade solid dielectric, thus eliminating any gaps in which electrons can accelerate.
Critical technical, market, and policy issues relevant to future large broadband switched satellite networks are summarized. Our market projections for the period 1980 to 2000 are compared. Clusters of switched satellites, in lieu of large platforms, etc., are shown to have significant advantages. Analysis of an optimum terrestrial network architecture suggests the proper densities of ground stations and that link reliabilities 99.99% may entail less than a 10% cost premium for diversity protection at 20/30 GHz. These analyses suggest that system costs increase as the 0.6 power of traffic. Cost estimates for nominal 20/30 GHz satellite and ground facilities suggest optimum system configurations might employ satellites with 285 beams, multiple TDMA bands each carrying 256 Mbps, and 16 ft ground station antennas. A nominal development program is outlined.
The thermal and mechanical characteristics which will ultimately limit the performance of Nd:BeL at high average power levels were investigated. The output beam characteristics (pulse width, peak power, beam dimensions and collimation) were determined at high repetition rates for both Nd:BeL and Nd:YAG. The output of Nd:BeL was shown to exceed that of Nd:YAG by a factor of 2.7 at low Q-switched repetition rates (1 Hz). This result follows from the smaller stimulated emission cross section of x-axis Nb:BeL compared to that of NdYAG by the same factor. At high repetition rates (10 Hz) the output of Nd:Bel falls to a level of three-fifths of its low repetition rate value while under similar tests the output of Nd:YAG remains essentially constant. A comparison of the measured values of the elasto-optic coefficients, the dn/dT values and the linear expansion coefficients for BeL and YAG failed to provide an explanation for the performance of BeL; however, thermal lensing was observed in Nd:BeL. Results imply that the output of a high repetition rate Q-switched Nd:BeL laser (high thermal loading) could be dramatically increased by utilization of a resonator design to compensate for the thermal lensing effects.
A 19 GHz waveguide latching switch with a bandwidth of 1400 MHz and an exceptionally low insertion loss of 0.25 dB was demonstrated. The RF and driver ferrites are separate structures and can be optimized individually. This analysis for each structure is separately detailed. Basically, the RF section features a dual turnstile junction. The circulator consists of a dielectric tube which contains two ferrite rods, and a dielectric spacer separating the ferrite parts along the center of symmetry of the waveguide to form two turnstiles. This subassembly is indexed and locked in the center of symmetry of a uniform junction of three waveguides by the metallic transformers installed in the top and bottom walls of the housing. The switching junction and its actuating circuitry met all RF performance objectives and all shock and vibration requirements with no physical damage or performance degradation. It exceeds thermal requirements by operating over a 100 C temperature range (-44 C to +56 C) and has a high power handling capability allowing up to 100 W of CW input power.
Both p channel and n channel enhancement mode MOSFETs can be readily switched on and off at temperatures as low as 2.8 K so that switch sampled readout of a VLWIR Ge:Ga focal plane is electronically possible. Noise levels as low as 100 rms electrons per sample (independent of sample rate) can be achieved using existing p channel MOSFETs, at overall rates up to 30,000 samples/second per multiplexed channel (e.g., 32 detectors at a rate of almost 1,000 frames/second). Run of the mill devices, including very low power dissipation n channel FETs would still permit noise levels of the order of 500 electrons/sample.
In a push-pull converter, switching transistors are protected from peak power stresses by a separate snubber circuit in parallel with each comprising a capacitor and an inductor in series, and a diode in parallel with the inductor. The diode is connected to conduct current of the same polarity as the base-emitter juction of the transistor so that energy stored in the capacitor while the transistor is switched off, to protect it against peak power stress, discharges through the inductor when the transistor is turned on, and after the capacitor is discharges through the diode. To return this energy to the power supply, or to utilize this energy in some external circuit, the inductor may be replaced by a transformer having its secondary winding connected to the power supply or to the external circuit.
The repetitive Q-switched performances of an x-axis neodymium-doped lanthanum beryllate (Nd:BeL) laser and the current state-of-the-art Nd:YAG laser are compared. Output beam pulse energy, peak power, pulse width, beam divergence and mode pattern were measured for a gold-plated single elliptical cavity using a xenon flashlamp for excitation. The output of the Nd:BeL laser is found to exceed that of the Nd:YAG laser by a factor of two at Q-switched repetition rates on the order of 1 Hz, due to the smaller stimulated emission cross section of the x-axis Nd:BeL. When the repetition rate is increased to 10 Hz, the Nd:BeL output falls to a level three fifths lower than its low-rate value, while Nd:YAG output remained essentially constant. Analysis reveals that the decrease in output may be attributed to a depolarization loss due to thermally induced stress birefringence.
A new technique using the mathematical nonlinear programming ALAG is proposed to facilitate design optimizations of switching power converters. This computer-aided approach provides a minimum weight (or loss) design down to the details of component level and concurrently satisfies all related power-circuit performance requirements. It also provides such design insights as tradeoffs between power loss and system weight as the switching frequency is increased.
A satellite-switched, multi-beam system using N sub T transponders is considered, in which a large number of ground stations are distributed over N sub Z (not less than N sub T) zones. Each station transmits information digitally in a packet-switched time division multiple access mode, with all packets of fixed length. A demand-assigned packet scheduling procedure is described which achieves high throughput with relatively low average time delay, adapts easily to fluctuating demand from individual earth stations, and automatically schedules transmissions to avoid adjacent zone interference.
Improved switching transistors handle 400-A peak currents and up to 1,200 V. Using large diameter silicon wafers with twice effective area as D60T, form basis for D7 family of power switching transistors. Package includes npn wafer, emitter preform, and base-contact insert. Applications are: 25to 50-kilowatt high-frequency dc/dc inverters, VSCF converters, and motor controllers for electrical vehicles.
Coupled-inductor polyphase regulator has better efficiency and lower inductor losses. Improved two-phase switching regulator employs negative coupling between inductors to achieve better power-to-weight ratio while reducing peak switching currents and inductor losses. Improvement of about 35 percent using new technique.