Continued from the preceding article on the design of a microswitch based on analysis of actiorr. and deformation energy of component parts, another type of microswitch, for which the calculation of deformation energy appeared very difficult, is similarly dealt with.
A high voltage pulse transformer with a magnetic core which is biased by d.c. current in the direction opposite to the pulse magnetization has been developed. Pulse characteristics of core materials are studied in two cases of with and without d.c. bias. A pulse transformer for. a klystron amplifier with a step-up ratio of 13:1, the pulse duration of 4 μsec and the secondary peak voltage of 150kV has been constructed. Rise time of 0.3 μsec and linear droop of less than 1.5%/μsec are obtained.
Magnetic field used in usual mass spectrometers is of so-called sector or Nier type with straight boundaries. In this paper, a circular magnetic field, which was originated by Bainbridge and has hitherto been disregarded, is taken up to apply it to multi-dispertion system with ion mirrors, simplifying the construction and adjustment of the spectrometer. Analysis of ion trajectory in this magnetic field is given. In a symmetrical circular magnetic field, spherical aberration by a right angle deflection of ion beam becomes 4Rα3 where 2α is the aperture angle of the ion beam at its source and R the radius of trajectory. Rα2 being the spherical aberration in a symmetrical sector field, the advantage of the circular field over the sector field is obvious; a significant improvement upon resolving power and sensitivity of mass spectrometer is therefore expected. by the use of circular magnetic field. Introducing the concept of “index of performance”, the author estimates the increase of the index at approximately 100% under ordinary working conditions. On the assumption that, in double-dispersion system-the simplest case of multi-dispersion system-with one magnetic field and one ion mirror, the ion beam to the mirror and that reflected from the mirrror are correctly on the same route, dispersion distance and the beam width at the collector are evaluated. Generally, chromatic and spherical aberrations increase in proportion to dispersion; in double dispersion system, they become roughly twice of those in single dispersion system while the width of the image of ion souce slit remains the same as that of the slit. Therefore, the resolving power now defind as the ratio of dispersion distance to twice of beam width, which is different from customarily accepted interpretation, do increase. Especially, the third-order spherical aberration, as well as the second-order aberration, vanishes in a circular field by a right angle deflection, and the resolving power becomes independent of the aperture angle of ion beam at its source. This means that the sensitivity of a mass spectrometer can be increased at will without lowering the resolving power.
(1) The relation between electric current and applied voltage is investigated on brominated graphite (residue compound) specimens of varied bromine atom percentages. This relation is found linear along two directions, parallel and perpendicular to extrusion axis of the specimens. Rectifi-cation effect is not observed. (2) X-ray diffraction patterns show that the lattice constant c increases linearly and a decreases slightly with increase of bromine atom percentage; up to 1.09%, -Δa/a is proportional to Δc/c with the ratio of 1/5. (3) The Hall coefficient of the residue compound at various stages of bromination is measured in the temperature range between 295°K and 105°K. The carrier density and the Hall mobility are calculated from the Hall coefficient. The sign of the Hall coefficient is negative up 3.0 atomic percent of bromine, beyond which it is positive; with further addition of bromine, value of the coefficient increases. Magnetic field dependence of the coefficient is inconsiderable. It is to be noted that, in the region of atomic percentage of bromine from 0.30 to 0.95, carrier density diminishes but electric conductivity increases with increase of the percentage.
On compensated high-resistive InSb crystals, very low values of carrier mobility are found at liquid nitrogen temperature, which is ascribed to irregular distribution of thermally unequilibrated lattice imperfection caused by innumerable crystal dislocations. Such lattice imperfection is also observable when InSb crystal is subjected to plastic deformation at 450°_??_480°C. When InSb crystal is bent, N-type becomes P-type, but P-type remains as P-type. Mobility of the converted P-type InSb becomes very small, whereas mobility of P-type InSb, once subjected to bending, is not much different from original. In photoconductivity measurement, the threshold wavelength “λ_??_” of uncompensated high-purity N-type InSb, carrier concentration of which is 5.1×1014cm-3, is found to be 5.51μ. On the other hand, λ_??_ of compensated P-type InSb with zinc accepter and that with dislocation accepter are 5.51μ and 5.75μ respectively. As for photosensitivity, InSb with dislocation accepter is found more sensitive than InSb with zinc accepter at liquid nitrogen temperature.
Dislocation of silicon single crystals grown by the floating zone method is examined by the copper decoration technique and X-ray Borrmann's method. The crystals used are those made by the authors, by du Pont Co. and by Siemens Co. Ltd. All these crystals have similar patterns of dislocation with dislocation densities of (1_??_4)×104/cm2. Interface between solid and melt during the crystal growth is observed by p-n junction method. Relations of the shape of interface to growth condition and crystal imperfection are investigated by varying the growth condition. Temperature distribution in crystal ingot is dealt with as a problem of heat conduction in a cylinder, and the calculation is carried out by Liebmann's method with the aid of an IBM 650. In the calculation, heat dissipation at the surface is assumed to follow Stefan-Boltzmann's law and temperature dependence of thermal conductivity of the crystal material is taken into consideration. Agreement between calculated and observed results is fairly satisfactory.
Singnal and noise characteristics of high impedance-type and low impedance-type lead sulphide photoconductive cells are obtained by experiment, and concentration and mobility of majority carriers (holes) are calculated from obtained data. Low impedance cell shows good signal-to-noise ratio and signal response at low bias voltage (below about 10 volt). Its current noise and non-current noise are both fairly low in comparison with those of high impedance cell. Generation-recombination noise is predominant in current noise of cells of both types. Low impedance cell has smaller time constant and larger carrier mobility than high impedance cell. This means that PbS film of low impedance cell has a better crystal disposition than that of high impedance cell. Hence, the annealing in the process of preparing low impedance cell is presumed to regularize the crystal disposition of PbS film and improve the cell characteristics if the state of oxidation is adequately maintained.
A new type gas flow proportional counter for detecting ultra-soft X-rays from low elements has been devised. Construction and performance of this counter are given in detail. Pulse height distribution curves of output by carbon, aluminium and magnesium K X-ray emissions show that the resolution of the counter for X-ray energy is almost of its theoretical value and that the pulse height voltage is proportional to the energy of X-ray radiation. For detection of X-rays of the wavelengths longer than 20Å with a proportional counter, absorption of window material is a matter of serious concern. Polycarbonate is used; it is easily made into thin films: thickness of a few thousand Å gives the film enough tensile strength to withstand pressure difference between inside and outside of the counter and is sufficiently non-permeable to gas. At 5000 Å of thickness, the film absorbes 50% of 44 Å carbon K X-ray emission by cal-culation. Analytical results are given on carbon on the surface of iron during casting observed with an X-ray scanning microanalyzer provided with the counter described above.
Vertical concentration distribution of impurities in ingot grown by resistance heating horizontal zone melting method is investigated. For the analysis of experimental data, Abe's model-natural convection due to temperature gradient perpendicular to vertical solid-melt interface is the most important factor that determines effective distribution coefficient of impurities; expressbile by Saunders' first approximation-is revised as follows. The rate of natural convective flow over the upper half of solid-melt interface in molten zone can be calculated by Saunders' theory, but over the lower half, the flow rate is smaller than the value expected from his theory owing to the effect of bottom; it is assumed a priori that the flow rate over the lower half of interface can be expressed symmetrically to that over the upper half of interface. Measurement of vertical resistivity distribution is made on <111> grown Ge single crystal doped with Li, As, Sb, Bi, Zn, Ga, In, and TI by zone levelling for which a travelling resistance furnace is used. According to the revised model, if the distribution coefficient is smaller than unity, the impurity concentration in lower part of the ingot becomes larger than in upper part when the growth rate becomes larger, the temperature difference in molten zone smaller and the diffusion constant of impurity in melt smaller, and the results obtained experimentally can be explained more or less numerically if we assume DAs≈DSb≈DBI≈DIn≈DTI≈6×10-5cm2/sec and DGa≈1.4×10-4cm2/sec. The revised model explains also why the forced convection effected by the application of D. C. current and static magnetic field method or travelling magnetic field method causes homogenization of resistivity in vertical derection.