VOL. 14 No. 5, pp. 117·152 NOVEMBER 1952 Philips Technical Review DEALING ~TH TEC~~CAL PROBLEMS RELATING TO THE PRODUCTS, PROCESSES AND INVESTIGATIONS OF THE PHILIPS INDUSTRIES
EDITED BY THE RESEARCH LABORATORY OF N.V. PHILIPS' GLOEILAMPENFABRIEKEN. EINDHOVEN.NETHERLANDS
ELECTRONIC TUBES
A synopsis hy J. L. H. JONKER 621.385
The following article covers the main points of ctn address by Prof Dr. J. L. H. Jonker on his formol acceptance ofthe office of professor at the Technical Un iuersiiy at. Delft on 19th March 1952. Although the extent of his subject did not admit of more tluin a brief review, this "bird's-eye view", which brings mall)' ioulelv separated aspects into (/ common focus,
has an appeal of its 01011.
Nothing hut the greatest admiration can be felt It was roughly at the turn of the century that by anyone tracing the rapid development of elcc- Hertz, Sir OliveI' Lodge and Marconi carried tronies, that is, electronic tubes and their applica- out the experiments that led to the successful tions, which has sprung from the phenomenal introduetion of telegraphy without wires by means growth of radio in the first half of the twentieth of high-frequency oscillations. century. In the short span of a few decades radio has 118 PHILIPS TECHNICAL REVIE W VOL. 14, No. 5 undergone almost unbelievable technical develop- "so outstanding in its consequences that it almost ment and improvement, and it has achieved a ranks with the greatest inventions of all time". revolution in the interchange of ideas between This invention is the triode of Lee de Forest. individuals and peoples alike. It has come to assume The advent of the triode such importance in the lives of almost everyone that throughout the whole evolution of mankind In seeking for a better detector of radio signals there have been few technical developments to Lee de Forest 2) in 1907 introduced a grid between equal it. the cathode and anode of the existing diode devised by Sir John Ambrose Fleming, thus obtaining The early days of radio a means of controlling the flow of electrons between This development of radio, which took the world by storm, founds its origin in the experiments al- ready mentioned. Channels were established across the oceans, first telegraphically and later by tele- phone, up to the point where radio contact now covers the whole glo~e. In the early years, technical knowledge and literature on the subject were in the nature of things very limited. Because the individual experimenter was, however, in very Juany instances able to achieve valuable results and improvements using only simple and often home-made equipment, many felt themselves drawn towards this sphere and this is undoubtedly one of the causes of the very rapid advances made. Those who were privileged to witness at close quartcrs some part of this inspmug early period may look back on those exciting days with some rcgret, comparing them with conditions as they are today. now that improvement and research into specific problems are in thc hands of a corps of specialists having the most ingenious laboratory equipment at their disposal. As a result of the work of these specialists, Fi~. 1. The ori~illal triode C'audioll") of I,ee dl' Forest. literature on the subject of radio would now con- stitute a whole library in itself, whereas, on the other hand, thc ceaseless flow of publications cathode and anode without the cousurnption of auy presents to the expert the problem of keeping in energy (figs 1 and 2). Years elapsed before the touch with everything that is of interest in his operation and possibilities of this new invention work. Here, too, a too copious emission results in were fully appreeiated, in consequence of which so great a space charge that the ultimate object is De Forest became so short of funds that he had to defeated. In the United States this urgent problem let his European patents lapse! In 1911 Robert has recently given rise to an investigation into the von Lieben gave a demonstration of the ampli- statistical distribution of such publications among fication properties of the triode at the Berliner the various periodicals; the result of this investiga- Physikalische Gesellsehaft. In 1913different research tion has been reported - in another publication! - workers simultaneaously invented the "feed-back", in the periodical containing the largest number of which permitted the triode to he used as a very such publications 1). effective g en er a t 0 r of high-frequency oscillations. It is because of the initial circumstances already The last-mentioned invention gave rise to the outlined that history can point to one specific longest case of patents litigation in this field in invention of one man, as being as it were the lever history ~), a contest which was not settled until which released radio from its trammels. This was 1934, in favour of De Forest. the start of the subsequent fantastic growth and The invention of the triode, which marked the was, in the words of the Nobel-prize winner Ra b i: commencement of the electronic era, gave to radio - ,.,., -,-
NOVEMBER 1952 ELECTRONIC TUBES 119
at once a better detector, the long sought amplifier reproduction and the oscilloscope. In the years and an oscillator. These various possibilities provid- that. followed, research penetrated the realms of ed the stimulus for a careful study' of these tubes ultra-short waves, atmospheric interference, fre- and for such modifications as would ensure optimum quency modulation and television, apart from in- results in any function in transmitting and receiving numerable industrial applications of electronics. In equipment, and this has in turn produced the Ill- each case the electronic tube was the key that nu~erable varieties of electronic tubes which we opened the door to the new sphere, and .often,- know today." in new forms, to entirely new possibilities. This r .....fuN '70.~
-,I when ~d oondOJ'OOr is ~L OTer the dUcling member from ....-'"z; ·ltAIHO TI:I":PIl(l:\t: A ClJl'IIQ!tATION SEW "ORK .~R~~~~~\~:~:~~;~~'~':~~:~N~::~~"~ro , or 8Ounu.. produced therein under the Mm~ eon- chnrg"d. diLiona wllen Mid eondenser ia not employed. 'P~OK TIU.JI:Oa·~PBT. ;. An oscillation d, h will be understood thaL Lh. drcuil ar- P\'IiCUa.'4>t1 veSMI incl. l ra.ngemenls herein described with reference J:a8I.'O\l8 ... I7O,ua. 'lpec:tGoaUoa of 1A\t4.n ~&~t.. ducting me-di Pa&.ea.M4 hbo 1',1001 to the part.icular (Ofins ol audion herein dis- ruetubera inclosed the l":kalJ ... 1.. J..... 1l 211.1107. 8.,t&l I,. nUI2. eloeed lDay wiUI acJvänla~e also be employed tiun eircuit, a circuit c with v.riuu" other l\'IIt":J uIalIdion. • ol tl8ÏtI oscäleticn cie. To .11 tI\\om il may con",,",: Iclt.iw: . end bruughL uut to the temunal 3. Inter- u 1II('1II1wnt, a
Ir. JI..~(ill·;': frllill Ih" I'l'iIlηi't!t·!'O Hr lily inn'ntion, III hC':tt wuuld I",,·nlerely lA.'Ol~ll\'p, I du nul trod ... 10. An U8Cllldli,IO I lu t lu- dr .. " m;.ro., ,,'I:,:un' I' n'IITI':-4·II.ld iu CI"('11i Il nf!I'I'rlKMry herefu to tlltfor into. de- 10 4. An oecillafiun tlt\tet'wr cumprbiing an evacuatvd \f'~,1 I" 11738 Fig. 2. Extracts from the United States patent granted to Lee de Forest, covering the invention of his triode. The "grid-shaped member" ~s mentioned in the third claim.
Rapid development applies equally to the latest branches of electronics, During and after the first world war we have seen developed during the last war and subsequently the efforts that were directed towards producing perfected, e.g. radar, sonar and other aids to navi- electronic tubes of high and very high power for gatiou, radio relay stations and computing equip- the establishment of world-wide radio communica- ment. tion. Then came a remarkable offshoot in the field Electronics in general is much indebted to the of electronics: the radio broadcast, commenced second world war, that black page in the history with noble idealism, but later to be misused for of mankind, since technical development was urged commercial advertising and political propaganda. forward by the exigencies of the war without being Radio broadcasting has in many parts of the world hamperedby economic and commercial factors. Many called into existence enormous electronic• industries of the results obtained in this way have since proved whose annual output is now estimated at some to be of great scientific, technical and commercial 500 million tubes and a good 30 million receivers, "interest and have been adopted or further developed the turnover for 1952 heing estimated at many for peaceful occupations; '.moreover, new techniques gigadollars. In this way a new sphere of employ- are continually being a(tde~' to the already wide ment has been created for hundreds of thousands field of electronics, such being for example radio in the manufacture and application of electronic astronomy,' colour television and, microwave spec- tubes. . ' troscopy. The very extensive scope of application The young but rapidly expanding industries with of electronic tubes and the consequent diversity . their many specialists soon found more scope in of their form may be illustrated with reference to other directions, to mention only sound amplifi- some of the extremes a~ong the dimensions Ill- cation systems, soundfilm, improved gramophone volved. 120 PHILlPS TECHNICAL REVIEW VOL. 14, No. 5
The diversity of present-day eleetronic tubes recClvmg tube, consisting In the attainment of It is possible by means of electrometer trio des smaller electrodes and closer spacing between the to measure currents of 10-15 amperes, whilst in electrodes, that have resulted in an over-all reduc- transmitting tubes and gas-filled rectifiers, current tion in the size of the tube (fig. 3). The inherent peaks of 100 amperes or more occur. Using elec- advantages are, amongst others, lower current con- tronie tubes we can measure alternating voltages sumption and better short-wave characteristics; of 10-6 volts (it should he recognized here that manufacturing costs have dropped owing to more in that case out of a possible 1018 available electrons rapid production, lower consumption of materials, in an amplifying tube, only about 1010 are control- smaller stores and reduced transport costs. The led). On the other hand, an installation is being receivers in which the tubes are used could be built by the Stanford University in California (a made more compact in consequence. "linear accelerator") which, with the aid of enor- Intensive research has taken place into the many mous transmitting tubes (klystrons), operated at physical phenomena occurring in electronic tubes 300000 V; will accelerate electrons to the equivalent that may be useful or detrimental, such as thermal, of a thousand million volts 4). Electronic tubes can secondary and photo-emission (fig. 4), fluorescence, he made to operate within a range of from 0 to 1011 gas discharge, space charge, potential fields, elec- cycles per second. The weights of conventional types tron paths, noise and, in short-wave operation, of tube may differ one from the other by a factor of transit-time effects and damping. Investigations as much as 100000. into these effects have not by any means ended; Within these extreme limits there is an enormous on the contrary, efforts are still being made, based variety of design among electronic tubes; a cautious on a growing knowledge of these often highly estimate places thc number of different commercial complex phenomena, to improve the performance types at about 20000. of the tubes. It was, for example, possible to improve It is not within our scope to give even the briefest the efficiency of pentodes considerabely by taking review of all these different designs. We shall into account in the design the paths of the electrons confine ourselves to a short discussion of some of through the potential fields of the grids 5). In most the more important applications only. cases, however, a computation of these potential fields and the paths of the electrons within them Receiving tubes would take up far too much time, or would prove The must important field of application of tubes to be impossible or possible only under very re- numerically is without a doubt to be found in stricted conditions. By means of models of the radio receivers. In the last decades there electrode system in the "electrolytic tank" or on has heen a refinement in the design of the a rubber sheet (figs 5 and (5), a better idea is obtained
Fig. 3. In the course of time the dimensions of radio receiving tubes have been reduced again and again. All the tubes shown are pentodes. NOVEMBER 1952 ELECTRONIC TUBES 121
Photograph Waiter Nürnbcrg
Fig. 4.. Apparatus employed for the measurement of secondary emission of solids. For a description see J. L. H. J onkel', The angular distribution of the secondary electrons of nickel, Philips Res. Rep. 6, 372-387, 1951 (No. 5). of what occurs inside the tube and thus of tubes will operate effectively at wavelengths down the means of simplifying the computations 6). to a few metres 8). The special form of construction In the course of time endeavours to secure im- necessary for tuhes intended to work at still shorter proved tube performance in the receiver have on wavelengths will he considered under a separate many occasions shown the desirability of including heading. more than one grid between cathode and anode, Whereas the configurations of the electrodes in and in this way the well-known tetrodes, pentodes, tubes for the various functions in receivers have hexodes, pentagrids, octodes, having respectively hecome more or less stabilized in recent years, 2, 3,4,5 and 6 grids, originated. The record number the same cannot he said of tuhes for television of grids so far is found in the enneode (7 grids) r ecept ion. The short wavelengths, th e wide employed as a phase detector for F.M. signals 7). frequency band and the sav-tooth voltages required Small dimensions and. a high degree of accuracy for the motion of the cathode ray over the screen in the arrangement of the electrodes and their impose other requirements on the tubes than those leading-in wires ensure that these broadcast necessary for ordinary broadcast reception. These 122 PHILIPS TECHNICAL REVIEW VOL. 14, No. 5
which the tube materials are able to sustain deter- mines the limit of utilization and, if such be possible, technology here plays an even more important part than in receiving tubes. Amongst classical types, i.e. the triode, tetrode and pentode, the volume of the tube for a given output power has in recent years been reduced hy a factor of from 5 to 10 (fig. 7) by employing high-melting-point materials with good radiation properties for the electrodes, non-volatile getters and new types of glass having improved thermal and electrical characteristics, as Fig. S. This deformed rubber sheet reproduces the potential field in the cross section of a plane triode having 5 grid rods, for given values of the grid and anode potentials. Steel balls rolling over the sheet describe the same pa ths as electrons in the potential field. requirements are again best met by tubes which are specially designed for the purpose 9). A similiar situation will arise in tbe future in the case of colour- television receivers, and we have every justification for anticipating that new designs will be evolved which will render the receivers more sensitive and less complicated. Efforts to reduce the number of tubes per receiver hy introducing special typcs would at first sight appear to constitute a laborious, suicidal policy on thc part of the manufacturer. This is by no means thc case, for simpler and less expensive receivers can be sold in larger quantitics. Apart from that, it is difficult to keep in check improvemcnts allel sim plificat.ions of techniques.
Tra n.smi /ûng tubes In transmitting tubes, thc mam function of which is to supply considerable power with a high degree of efficiency, the maximum temperature Fig. 7. Evolution of' the transmitting tube with the years. All three types deliver about] 000 W; the latest of these i, J4,5 cm in h eig h t.
well as by making a special study ofthe dimension- ing of the electrodes and their leading-in wires 10).
The reduction III inter-electrode capacitances brought ahout by thus curtailing the size ofthe tubes is particularly important at the very short wave- lengths, where low capacitance is a conditio sine qua non. The maximum permissihle loading of such small electrodes, however, sets a limit to the output power. When it is not possible with given electrode dimensions to dissipate enough of the heat hy radiation, forced cooling with air or water is employed (fig. 8), for which purpose highly
Fig. 6. Example of path of ball on the rubber sheet as photo- efficient cooling systems have beeu designed in graphed in intermittent light. The photograph shows that which the most advantageous turhulence of the under certain conditions the electron can undergo such deflection between the grid wires that it is flung beyond the coolant is ensured 11). In the future, especially 'pnrr between the elertrodes. among short-wave transmitting tubes, further NOV EMBEH 1952 ELECTIWNIC 'J'UBES
The ordinary recervmg tube is not effective at deeimetric and shorter waves, when the frequency is increased to such an extent, thc gain per stage
drops below unity. The causes of this, VIZ. electrical losses, radiation and transit-time cflocts, are in part ascribable to the characteristics of the circuit at such wavelengths; in part, too, they arc of electronic origin 13). Here, other forms of tuning devices, as well as tubes of different design are needed to ensure success; for example, in the centimetric and millimetric wavelengths, the cir cui t consists of transmission lines, coaxial cables, wave-guides aud cavity or rod-shaped resonators 14), the radiation and I2R losses of which can be made appreciably lower than those of the conventional types of coil employed for the longer wavelengths. It has been found possible to extend the working range of the tubes towards the shorter wavelengths, firstly by reducing the size and spacing of the electredes so as to cut down capacitances and transit-time effects and, secondly, by making the leading-in wires for the electrodes function as part of the resonant circuits. The latter step has led to the use of disc-shaped or annular electrode connections in short-wave tubes; the metal discs, being thus part
Fig. 8. Water-cooled transmitting tube for 100 kW output, of the resonant circuit, are sealed laterally through with its anode water-cooling jacket. the wall of the glass envelope, so that the part inside can serve as support for an electrode (jig. 9). improvements In consequence of continued tech- As to the reduction in the electrode spacing, it nological research, as also further refinements In has in some cases been found possible to cut down the electrode system, are to he anticipated. the space between the control grid and cathode to
Tubes for ultra-short waves The urge towards the new and unknown has prompted many experimenters to investigate shorter and still shorter wavelengths. For the original radio links over great distances the long-wave range of 1000-20000 metres was employed, but techniques have in the course of time tended more and more towards shorter wavelengths for such purposes. As to this, we are much indebted to amateurs all over the world for the part they played in this pioneer work. At present normal communication transmitters, amongst which those intended for broadcasting, operate at wavelengths ranging from 25 kilometres to about 1 metre, whilst for normal television wavelengths of several metres and for colour television decimetrie waves are used. Centimetric waves are now used for radio relay stations and for radar; millimetric waves serve physical science in the study of the characteristics· of molecules and atomic nuclei by means of the Fig. 9. "Disc-seal triode" type EC 56. Grid and anode are carried by metal discs sealed laterally into the glass wall; tlw absorption spectroscopy of gases 12). outer part of the discs forms part of the resonant circuit. 124 PHILlPS TECHNICAL REVIE\V VOL. 14, No. s as little as a few hundredths of a millimetre 15). emission of the ordinary oxide-coated cathode can So that the controlling action shall not be lost in be considerably increased 16). consequence of this, the mesh of the grid has to In order to lengthen the life when the specific he unusually fine; 'to this end taut tungsten wires emission is increased, special cathodes have been 10 microns in thickness are used, this being many developed, such as the L cathode in which the same times thinner than human hair (fig. 10). Transit alkaline earth oxides are used as for the conventional times in an oscillator can be reduced still further by oxide-coated cathode, but applied behind a porous the use of higher voltages, which, in order not to layer of sintered tungsten 17). overload the electrodes, are applied in pulses; the If high power is required at very short wave- electrodes thus have time to cool between two pulses. lengths, it is necessary to resort to electrode systems whose dimensions are of the same order as the wave- lengths at which these systems should operate; this, now, is the case in tubes whose action is based on a utilization of transit-time effects, as in magnetrons which during the second world war played such an important part in radar for locating submarines and aircraft, as also in klystrons, travelling-wave tubes and similiar systems 18). In such tubes the transit time of the electrons is of the same order as the periodic time of the alternating current; thus the electrons are able to impart energy to the high-fre- quency field which is thereby increased so that oscillations are produced. Although such systems do not excel by reason of their efficiency, they have the advantage that they are large enough with respect to the short wavelength not only to dissipate the heat generated, but to be manufactured without serious mechanical difficulties. For gencrating waves of 3 and 10 cm, klystrons ami magnetrons have been made that supply a pcak output of rcspectively 10 and 5 megawatts, with efficiencies of 30% and 60%. On the other hand, a 3 cm magnctron has been manufactured for a peak output of 100 W which weighs not more than 1 kg including the permanent magnet (fig.ll). Fig. 10. Micro-photograph of the grid incorporated in the The shortest wavelength for which magnetrons triode depicted in fig. 9. The thickness of the grid wires is roughly 10 I).. A human hair is shown across the wires for can be manufactured is in the region of a few milli- comparison. Magnifi('ation 10 x. metres, the efficiency being rather low. It has also proved possible to obtain a small amount of power, Clearly there is a practical limit beyond which as higher harmonic of the fundamental wave, at a it is not possible to manufacture tubes with still wavelength of roughly 1 millimetre 19). smaller dirnensions; it would seem irnprohable, there- Klystrons have been made for a continuous fore, that we shall he able to get any further than oscillation at a wavelength of a few millimetres, wavelengths of some centimetres by this means. but, as the circuit losses are inversely proportional Ultimately the limited loading capacity of the elec- to the square root of the wavelength, the efficiency trodes, as well as the impracticability to fix elec- at such wavelengths is very low. trodes operating at elevated temperatures with These tubes are accordingly not suitable for extremely small spacings, will prove the deciding amplification purposes. The travelling-wave tube factors. Another limitation is found in the size of is much less susceptible to poor circuital properties the cathode, since emission, and therefore also the in this range of frequencies, as it operates without output power, decrease with the area of the a resonant circuit. cathode. Increased emission through higher specific In these tubes a beam of electrons, usually con- loading can be achieved only at the expense of the centrated by means of a magnetic field, is shot life. Under pulsed conditions, however, the specific in axial direction through an elongated wire helix, NOVEMBER 1952 ELECTRONIC TUBES the signal being applied to the cathode side of the Cathode-ray tubes helix. When the speed of propagation of the beam In cathode-ray tubes use is made of a narrow and that of the field of the travelling wave in the beam of high-speed electrons to produce a luminous axial direction are made almost identical, the signal image. This is achieved by deflecting the beam, or is amplified along the length of the helix, at the cathode ray, electrically, so that it travels over a expense of the beam. With greater losses in the surface coated with substances that emit light in helix it is only necessary to employ a longer helix consequence of the electron bombardment. Origin- for the same amount of gain. ally these tubes were employed in the cathode-ray oscilloscope. The coming of television has imposed new conditions on the quality, and mass producrion, too, has brought other considerations to bear. A distinction is made between two systems in tele- vision, viz. direct and indirect vision of the images. Until now the former has been the more widely used ofthe two systems, but this has led to the making of tubes with larger and larger screens, seeing that small images tend to tire the viewer and to limit the number of viewers per receiver. In order to reduce the weight of such large tubes, the heavy glass cone has heen replaced by a lighter one of metal 20) and the inconveniently largc dimensions have been cut down by introducing a rectangular screen to take the place of the round one (see also frontispiece), and shorter electrode systems for generating and deflecting the electron beam. It has also been found possible to shorten the cone by employing a narrower electron beam which is Fig. ll. Magnetron for a wavelength of 3 cm designed by capable of deflection through larger angles (90°) Philips Laboratories, Irvington, N.Y., U.S.A. This magnetron is not intended for radar, but for a radio beacon. Peak output (fig. 12). 100 \V at an anode voltage of 800 V, heater power 2 \V. Total It may well be asked whether it is desirable to weight including the permanent magnet: approx. 1 kg. The size may be seen from the inch rule placed in front of the mag- gu any further in this direction to produce still netron. (Photo by courtesy G. A. Espersen and B. Arf in, larger pictures. The presence in the living room of Tele-Tech. 10, No. 6, p. 50 and No. 7, p. 30, 1951.) a vacuum tube 75 cm in diameter, as recently produced in the United States, would seem to be On this principle tubes have been made that will anything but desirable, even though the danger of provide ample gain at a wavelength of only a few implosion due to the atmospheric pressure of centimetres, and the success achieved with these roughly 4 tons on the glass screen need not be tubes on such wavelengths has stimulated experi- considcred high in a properly designed tube. In ments with numerous similar systems, in which the order to withstand such forces, the screen has to be helix is replaced by rods, wires, a wave-guide, or very thick, and also more or less convex, which a second electron beam. means a certain amount of distortion for the viewer. There is still an enormous amount of scope for Lastly it may be said that a tube of such dimensions microwave experimentation, if only by reason of presents quite a problem for the set manufacturer the fact that among the many different methods in the design of a cabinet that will be suitable for of making a high-frequency alternating field and a installing in the living room. constant electric and magnetic field cooperate If large or very large pictures are demanded, the spacially with a beam of electrons, only few have indirect system is much to be preferred; a cathode- been investigated in tubes. More systematic research ray tube of quite small dimensions but very high than has hitherto taken place might yield remark- luminous intensity then provides an image that can able possibilities. be projected on to a flat screen by means of lenses In many laboratories all over the world efforts or a concave mirror 21). The high luminous inten- are being made along widely divergent lines to sity of the screen required in this case imposes very penetrate further into the interesting problem of stringent conditions on the luminescent substances generating and amplyfing microwaves. from the point of view of loading. 126 P1LlLIPS TECHNICAL HEV1EW VOL. JA, Nu. 5
Camera tubes for television per second by a beam of electrons, the necessary Developments in cam.era tubes for television signal being obtained by "measuring" the charges have led to magnificent and most interesting results. of the photo-cells along those lines. We have been given the iconoscope, the image icon- Means have been found of ensuring a very high oscope (fig. 13), theorthicon and the image orthicon, sensitivity to light in the image orthicon, namely associated with such names as Zworykin, Hose 10 times higher than that of the fastest photo- and I a m s 22). These tubes, which convert the graphic material; hence a serviceable image is ob- optical image into an electrical signal and which tained by the light of a candle, or by moonlight. rank amongst the most complex of electronic inven- This has been achieved, firstly, hy using for the tions, were the means of appreciably hastening for- scanning a beam of electrons having a comparatively ward the realization of practical television. The low velocity, thus eliminating the interfering technological problems to be solved were great. For secondary emission from the mosaic inherent in the
Fig. 12. Ca thod c-ru y tubes fur television. At the left a rectang u lar direct-vision tube with metal cone giving a picture 28 ern X 37 cm. Centre: tube for projection television in t he home (max. 1.0 m X 1.2 III appro x.), At the right: a tube for projection in halls (3 m X 4 m).
instance 111 the iconoscope the optical image is iconoscope and, secondly, by making effective use of translated into a pattern of electrical charges by sccondary emission by including in this already means of a screen or mosaic about 100 sq. cm complicatcd tubc a multi-stage electron multiplier. containing several milliards of minute photo-cclls, Thus, in order to ensure distortionless conversion each insulated from its neighbours, in a density of of thc optical image into electrical signals in these roughly 360000 per sq. mm. For cornparati vc tubes, a number of electronic mechanisms, each of purposes it may be pointed out that the concentra- which is already sufficiently complicated, must be tion of the rods and cones in the retina of the human madc to work: together in the correct manner, eye is at most 20000 per sq. mm, which means that which means that the manufactnre of this kind of the artificial retina of the iconoscope improves on tube is certainly no sinecure. As with other types nature to the extent of almost 20 times (fig. 14). of tube, endeavours are being made to render the In order that the image may be transmitted, it is carnera tube a more compact instrument: amongst scanned in lines at a speed of thousands of metres other things, this will admit of more convenient NOVEMBER 1952 ELECTRONIC TUBES 127
Fig. 13. Image iconoseope. At the near cnd of the tube wiIJ be seen the photo-cathode on which the scene to be televised is reproduced by means of a photographic lens. The large magnet coil shown a t this end of the tube produces the field required to focus the image from the photo-cathode on to the mica target, of which a part is just visible at the opposite, wide, end of the tube. At the far side the neck with electron gun and de- flection coils for producing t.h« electron he am which periodically scans the target. proportioning of th e optical system, increased In conclusion dcpth of focus and less expensive lenses. It is not unlikely that in the future higher sensitivity in this Only a few examples from the very extensive kind of tube will be sought in a wider use of photo- field of electronic tubes have been discussed here: conductivity in place of photo-emission, seeing other important applications, such as those of that the quantum efficiency of the former is in counter and "memory" tubes, gas-filled tubes, excess of unity. The phenomenon of induced con- switching tubes, tubes for physical research ductivity in insulators also has possibilities to offer or for industrial uses and many others have in the design of camera tubes 23). been omitted from our review. In most of these
Ct b
Fig. 1.4.. a) Photograph of part of the retina of the human eye magnified 750 times. (From S. L. Polyak, The Retina, Univ. Chicago Press, Chicago 1941). b) Part of the mosaic of an iconoscope, photographed with the electron microscope. Magnification 11 000 X. The part shown has the same size as that area of the retina enclosed within the smallmarked rectangle in photograph (a). 128 PHILIPS TECHNICAL REVIEW VOL. 14.,No. 5 categories exploratory ,~ork is still in full progress tubes are still invisible; even after half a century and new departures are of almost daily occurrence. of research and development; fresh possibilities are The limits of the possibilities offered by electronic continually being discovered.
BIBLIOGRAPHY ,The references addedto Prof. Jonker's address make no pretence at heing-a corn- prehensive review or even a representative cross seetion of existing Iiterature on the subject dealt with. In this selection the work of the Physical Laboratory at Eindhoven has naturally been placed somewhat in the foreground.
I) R. C. Coilc, Proc. I. R. E. 38, 1380-1384, 1950. 73-84,19'17 (cavity resonators). W. Opcchowski, Philips 2) U. S. Patent No. 879-532,applied for 29 Jan. 1907, granted techno Rev. 10, 13-25 and 46-54., 194.8 (wave-guides). 18 Febr. 1908. S. Mi l lm an, Proc. I. R. E. 39, 1035-1043, 1951 (rod" 3) Franklin Inst." State Pcnn. for the Prom. Mech, Arts type resonators). I Comm. Sci. Arts, Rep. No. 3087, 8 Jan. 1941, p. 3-4. 16) .T.A. Morton and R. M. Rydcr, Bell Syst. techno .T.~9, 4) M. Chodorow, E.L. Ginzton, I.N eilsen and S. Sonkin, 496-530, 1950. I. R. E. National Convention, New York, March 1950. IR) E. A. Coomes, J. appl, Phys. 17, 647-654., 1946; M. A. 5) J. L. H. Jonker, Wirel. Eng. 16, 274-286, 344-349, 1939; Pomerantz, Proc. I. R. E. 34, 903-910, 1946. A. .T. Heins van dcr Ven, Wirel. Eng. 16, 383-390, I,) H. J. Lemmcns, M. J. Jansen and R. Loosjes, Philips '1'14-4.52,1939. techno Rev. 11, 341-350, 1950. G) C. L. Fortescue and S. \V. Farnsworth. Proc. Amer. 18) On magnetrons sce: J. B. Fisk, H. D. Hagstrum and Inst. Ei. Eng. 32, 757-772, 1913 (principle of electrolytic P. L. Hn r t mn n, Bell Syst. techn.' J. 25, 167-348, 1946; tank). brief survey by J. Verweel, Philips tcchn. Rev.14, 44-58, G. Hepp, Philips techno Rev. 4, 223-230, 1939 (description 1952 (No. 2). of clectrolytic tank). ' On klystrons see, e.g., D. R. Hamilton, J. K, Knipp and P. H. J. A. Kleijnen, Philips techno Rev. 2, 338-345,1937 J. B. H. Kup er, Klystrons and microwave triodes, (method of rubber sheet). McGraw-Hill, New York 1948; also B. B. van lperen, For application and results of the methods see, i.a., Philips techno Rev. 13, 209-222, 1952 (No. 8). .T. L. II. .Tonker, Philips' techno Rev. 5, 131-140, 194.0; On travelling-wave tubes see: R, Kompfncr, Wirel. Eng. .T.L. H. .Tonker and .B. D. H. Tellegcn, Philips Res. 24, 255-266, 1947; .T. R. Pierce and L. M. Field, Proc. Rep. 1, 13-32, 1945; , I. R. E. 35,108-ll1, 1947; also B. B. van lpcren, Philips .T.L. H. Jonker; P_hilips Res. Rep. I, 331-338,)946; 4, techno Rev. 11, 221-231, 1950. 357-365, 1949; 6, 1-13, 1951. For other systems see, e.g., L. M. Field, Elcctronics 23, ') .T.L. H. .Tonker and.A . .T.W.:M. van Overbeek, Philips .Tan. 1950, 100-104. techno Rev. 11, 1-32i 1949. ID) .T. A. Klein, J. H. N. Loubser, A. H. Ncthercot Jr. 8) F. Pr a kk e, J. L. H. Jonker and :M. .T. O. Strutt, and C. H. Townes, Rev. sci. lnstr. 23, 78-82,1952 (No. 2). Wire!. Eng. 16, 224-230, "1939;G. Aima and F. Prakke, 20) H. P. Steier, .T. Kelar, C. Y. Lattimer and R. D. Philips techno Rev. ;8, 289-295, 1946. Faulkner, R. C. A. Rev. 10,43-58, 1949. D) R. ~L Cohen, R. C.,A. Rev. 12, 3-25, 1951. 21) P. M. van Alphen, H. Rinia, J. de Gier, G. .T.Siezen, 10) For a survey see, e.g., J. P. Hey boe r, Transmitting valves, F. Kerkhof and .T. Haantjes, Philips techno Rev. 10, revised by P. Zijlstra, Philips' Technical Library, 1951. 69-78, 97-104, 125-134, 307-317 and 364.-370, 1948/49. J1) H. de Brey and H. Rinia, Philips tcchn. Rev. 9, 172~178, 22) V. K. Zw or y ki n, Proc. I. R. E. 22, 16-32, 1934 (icono- 194.7(air cooling): M. .T.Snij d er s, Philips techno Rev. 10, scope); H. lams, G. A. Morton and V. K. Zworykin, 239-246, 1949 (water cooling). Proc. I. R. E.27, 541-547, 1939 (image iconoscope); A.Rose 12) Summarizing articles on this subject: W. Gordy, Rev. and H. lams, R. C. A. Rev. 4, 186-199, 1939 (orthicon); mod. Phys. 20, 668-717, 1948; D. 1(. Coles, Adv. in A. Rose, P. K. Weimer and H. B. Law, Proc. I. R. E. 34, Electronics 2, 299-362, 1950. 424-4.32, 1946 (image orthicon). 13) M. .T.O. Strutt and A. van der Ziel, Proc I. R. E. 26, See also P. Schagen, H. Bruining and J. C. Er an ck cn, 1011-1032, 1938. Philips techno Rcv. 13, 119-133, 1951 (No. 5). lol) Sec, e.g., C. G. A. von Li n d er n and G. de Vries, Philips 23) P. K. Weimcr, S. V. Forgue and R. R. Goodrich, tcchn. Rev. 6, ,217-224, 1941. (box-shaped resonators); 6, R. C. A. Rev. 12, 306-313, 1951 (photo-conductivity); 240-249, 1941 (transmission lines); 8, 149-160, 1946 (flat L. P'e n s a k. Phys. Rev. 75, 472-478, 1949 (induced cavity resonators). G. de Vries, Philips techno Rev.9, conductivity) .
. , NOVEMBER 1952 J 29
LIFE TESTS IN THE ELECTRON TUBE FACTORY
Photograph Waiter Nümberg