Reference / Paper · 1937
Bell Laboratories Record, July 1937 (The Crossbar Switch)
Bell's introduction to the crossbar switch: bars, select and hold magnets, fingers and contacts.
Reference / Paper · 1937
Bell's introduction to the crossbar switch: bars, select and hold magnets, fingers and contacts.
.h.c. J11!. 1'7 Íi; i F:rH ;..{il,'rc LAB ORATORI ES RECORD 1 v o 00 , THE CROSSBAR SWITCH 4- o J. N. REYNOLDS COAXIAL SYSTEM : CARRIER SUPPLY L. C. o / PETERSON GROUP TERMINALS A. G. JENSEN JULY 1937 Vol. XV No. 11 n. o t 171744 111-lIIJI'/J1JlfIJi IIJI11'110 BELL LABORATORIES RECORD Published Monthly by BELL TELEPHONE LABORATORIES, INC. PAUL B. FINDLEY, Managing Editor PHILIP C. JONES, flssociate Editor Board of Editorial Advisors E.W. ADAMS W. FONDILLER W. H. MARTIN H. A. AFFEL H. A. FREDERICK W. H. MATTHIES H. M. BASCOM B. C. BELLOWS O. M. GLUNT M. J. KELLY D. A. QUARLES G. B. THOMAS R. R. WILLIAMS JOHN MILLS SINGLE COPIES $0.25; subscriptions are accepted at $2.00 per year; foreign postage $o.6o extra per year. Subscriptions should be addressed to Bell Laboratories Record, 463 West Street, New York City In this Issue The Crossbar Switch 7. N. Reynolds 338 A Power Amplifier Tube for Ultra -High Frequencies fl. L. Samuel 344 Rectifier for Telephone Power Supply D. E. Trucksess 350 A High -Quality Headset for Monitoring F. S. Wolpert 353 Carrier for Coaxial Groups 357 L. C. Peterson Group Terminal for the Coaxial System 361 1. G. Yensen Volume 15- Number II-Iuly, 1937 HELL LAßORATORIES RECORD A Western Electric quartz plate unit photographed on a blueprint of crystal- controlled oscillator circuit a JULY, 1937 VOLUTE FIFTEEN- NUN113ER ELEVEN 44) q9q9szP The Crossbar Switch By J. N. REYNOLDS Apparatus Engineer THE problem of telephone apparatus used to make connections switching, or how best to connect any one telephone line to any other, has always been of fundamental importance to telephone engineers. It is an extremely complicated problem, however, and may be subdivided in various manners, depending on the aspects to be particularly stressed or the degree of detail with which it is to be studied. One convenient division is into systems problems and apparatus problems. The first group arises primarily because of 00 9 0 0 0 0 0 0 0 0 0 y.----0----0---0---¢--0- ¿01 o o o I 8¢ 0 O O 0 O 0 0 O O 7 d 0 0 0 0 0 0 0 0 o 63 0 0 0 0 0 0 0 0 0 5¢ O O O O O O O O O 4( 0 0 0 o o o o o o 3¢ o O o O o o O o 0 2 ¢ 0 0 0 0 0 o o o o o o o o o o o 0 0 2 3 4 5 6 7 8 9 0 1 i I 1 Fig. i -The step -by -step switch includes moo groups of terminals and one brush the very large number of lines that must be capable of being interconnected, and considers trunking schemes and circuits. The second is concerned with the actual switching 338 between lines and trunks. The two phases of the main problem are not unrelated, since the form of circuits and system employed may affect the type of apparatus required, and conversely the type of apparatus available affects to a considerable extent the type of system that must be provided. In the crossbar switch there is made available a distinctly different type of switch, and one that offers very definite advantages over previous types. Its most effective utilization will require a somewhat different system of trunking and different circuits, but neither of these latter aspects need be considered in pointing out the essential nature and advantages of the crossbar switch itself. It is necessary, however, to indicate the basic characteristics of the earlier forms of switching to illustrate the specific difference of the crossbar type. The type of switch that is used almost exclusively in manual telephone systems (using "switch" in the broad sense as any means of connecting one wire or circuit to another) is the plug and jack. A trunk or line is permanently connected to a jack, and another trunk or line is connected to a plug- either directly or through some other connecting device; and to make the desired connection, the operator picks up the plug, locates the jack of the line or trunk desired and pushes the plug into it. When mechanical methods of switching were developed, they followed the basic principle of ;7uly 1937 99 87 86 85 84 0 2 p 0 0 0 0 0 O Ó 0 110n 0 0110 0110 0 0 0 Y O 0 0 0 0 0 0 0 0 0 0 0 0 0 0-0-0 0110 000000 0 0 0 0 0 0 0 0 0 19 21 23 25 27 29 31 33 35 37 39 41 43 45 3 5 7 9 II 13 15 17 I 60 58 56 54 52 50 48 46 44 42 40 38 36 34 32 30 28 26 24 22 20 18 16 47 49 51 14 10 12 0 0 0 0 59 2 53 55 57 6 4 8 Fig. 2 -The panel unit consists of 6o multipled terminals in each of ioo rows, and a brush is provided for each of the 6o terminals of the multiple plug and jack, but the jacks were replaced by small metal terminals arranged in compact banks, and the plugs were replaced by brushes. These were made to slide along the terminals of the bank until they reached the terminal of the desired line, when a connection would be made. Two forms of machine switching have been widely used in this country: the step by-step, and the panel system. In the step -by-step system one brush is employed for each bank, and it moves both vertically and horizontally until the desired terminal is reached. With the panel system, the brushes move only vertically, but the banks are larger, and accommo- date a number of 9 0 8 O 0 0 0 vided for each such bank, and when not in use it rests below the bottom row of terminals at the extreme left of the bank. The terminals in the bank correspond to the jacks of the manual system, and the brush corresponds to the plug. To establish a connection the brush is moved: up by the action of a magnet, which lifts it one row for each operation; and then across the row horizontally by another magnet which moves it one contact in the horizontal direction for each operation. To connect the trunk associated with the brush of this bank to trunk 97, for example, the brush would be "stepped" up to the 9th row . 0 00 °O °o °0 o 0 0 0 0 o° o° 0° 0° 0 0 Ó o° 0 0 o 0 0 0 o° Ì P; o° ° o° o° 0 0 brushes operating over o o° o° 0 o 0° ° O° 00 00 o° Ó o° Ó 0° ° o 0o o 0 Ó o° parallel vertical paths. The arrangement of the step -by -step switch is illustrated in Figure 2 r. Terminals for one hundred lines or trunks 8 II 4 5 6 7 9 2 3 are arranged in a bank consisting of ten rows Fig. 3 -In the crossbar unit two sets of terminals are of ten sets of terminals mounted in place of the one set that is used in the step -by -step each. One brush is proand panel systems and sliding contacts are eliminated 0° O 7 / 1 7uly 1937 10 12 13 14 15 16 17 18 19 339 and then over to the 7th terminal. The banks of the panel system also, for the most part, have terminals for ioo lines, but instead of being arranged in a square array they are all mounted one above the other, and the banks are thus one hundred sets of terminals high. There are sixty sets of terminals in each horizontal row thirty on one side of the bank and thirty on the other. All the terminals in the same row, however, are connected together so that they represent only one line or trunk. Instead of one brush at the bottom, as with the step -by -step switch, there are sixty one for each column of terminals on each side as indicated in Figure 2. The brushes on one side are assigned even numbers and those on the other side, odd numbers. Here, as in the step -by -step system, the terminals in the bank correspond to jacks, and the brushes, to plugs. The brushes are driven upward at a uniform rate by a motor-driven friction drive at the bottom of the frame, which usually consists of five such banks, each with its own set of brushes. Both of these systems work very satisfactorily under conditions for which they are most suitable. There are two respects, however, in which improvement seemed possible. One is that with either system a comparatively complex mechanism is required to operate the brushes. The other is that a sliding contact is required, and the terminals and brushes must be of some durable metal to withstand the wear. Unfortunately, the harder metals do not have as low contact resistance as the softer precious metals such as silver, and are more subject to the formation of poorly conducting surfaces. The crossbar switch brings improvement in both of these conditions. It avoids sliding contacts and thus facilitates the use of precious metal contacts, and it accomplishes the required connections with a much simpler mechanism. It does this by employing an entirely different and much more direct method of switching. In the manual sys:. tem, and in both the step-by -step and panel (III systems, which resemI QO ble it in this respect, the members of one set II SELECTING of terminals are banked FINGER together, and the memrte bers of the other set, ACTUATING SPRING which are to be connected to those of the HOLDING first, are arranged to BAR be moved up into conSELECTING tact with them. In the BAR crossbar system all such necessity of motion is avoided by mounting a pair of contacts at each posiFig. 4- Simplified schematic of the selection elements of a tion in the bank, and crossbar switch by eliminating the - - /// I e,. 340 / 7u/y /937 ' 9. i !L °` ,!! P1416 ?dim'sis íal ". *I !!! ' ; ! 1i, ?' , ki 1'`i'1 it' . IG M. mi mikaiti grid mi mi rid ini piktlllia, . bi w! li ''1 lé`3i il ! " s O ti:áitvda1"i Te .,r..,..I fk'''G'l n, n, ! a ,. 1 111i! Fig. 5 -The crossbar switch has 5 selecting bars and thus Io horizontal rows of contact groups, and up to 20 holding bars and thus 20 rows of vertical contact groups brushes completely. The arrangement is as shown in Figure 3. One contact of each pair is multipled with the corresponding contacts of the other pairs in the same column, and the other contact of each pair is multipled with the corresponding contacts of the other pairs in the same row. The horizontal multipling corresponds to that of the panel system, while the vertical multipling is as though the brushes of the panel bank had been provided in multiple with as many brushes in each column as there were rows. Instead of moving up a brush, therefore, it is necessary only to close the contacts at the proper position in the bank to make the desired connection. The motion of the brush is avoided. The only movement required is that of a mechanical link to close the required set of contacts. How this is accomplished is indicated by the simplified diagrammatic sketch of Figure 4. Between each pair of horizontal rows is a bar running completely across the bank, which may be rotated a small amount in either direction around its axis by the action of two magnets and armatures at one end. Wires projecting inward toward the contacts are attached to 7uly 1937 these bars at each intersection with the vertical columns. With the horizontal, or selecting, bars in their midpositions these wires, or selecting fingers as they are called, lie between the two rows of contacts, but when the bar is rotated in one direction the fingers move up to lie across the backs of the contacts in the row above it, and when it is rotated in the other direction, the fingers are moved to lie across the backs of the contacts in the row below, as indicated by the dotted lines in Figure 4. Along each column of contacts is a vertical, or holding, bar which -when rotated by a magnet and armature at one end-moves a vertical bar inward to press against all the selecting fingers in that column. If none of the selecting bars are operated when the holding bar moves in, the fingers will merely be pushed down between the rows of contacts and no connection is made. If one of the selecting bars is operated, the fingers of that bar will lie across the backs of one row of contacts, and when the holding bar operates, the contact at the intersection of the selecting and holding bars that are operated will be moved into contact by the action of the holding bar 341 against the finger which, in turn, lies across the back of the contact spring. The holding bar remains operated during the period of the call, but the selecting bar returns to normal immediately after the holding bar has operated. When the selecting bar returns to the central position, all the fingers return with it except the one held by the holding bar, thus leaving the selecting bar free for another selection with a different holding bar. The fingers are small and readily flex over the small arc of rotation of the selecting bars. The actual appearance of a crossbar unit is shown in Figure 5, and in partially schematic form in Figure 6. There are five selecting bars, and thus ten horizontal rows of contacts; and there are twenty holding bars, and thus twenty vertical rows of contacts, although other numbers of holding bars may be used. The contacts themselves are similar to those of an ordinary relay and each contact in Figure 3 represents several contacts in the actual switch. Similarly each of the contact points indicated in Figures i and 2 really represents a group of contacts in the step-by -step or panel banks. In the crossbar system, moreover, twin contacts of precious metal are provided, thus giving greatly increased assurance that a good connec- tion will be made. The gain in simplicity of operation is very obvious. In the step -by -step system, for example, the upward motion of the brush is caused by one magnet operation for each row the brush passes over, and similarly for the horizontal motion. In the panel system upward motion is caused by operating a clutch at the bottom of the frame, and then the brush is driven upward at a uniform speed by a power drive. The upward motion is actually accomplished in two steps, separated by a slight pause. In the crossbar switch, however, only two magnet operations, one immediately following the other, are all that are required. The avoidance of sliding contacts in the crossbar system is equally obvious. The contacts are merely pressed OPERATING MAGNETS o o 0 0 SELECTING SELECTING BAR FINGERS -Mq HOLDING BAR Fig. 342 6- Partial perspective of the crossbar switch July 1937 together as in a relay when a connection is made, and no sliding in the ordinary sense occurs. By this adoption of a new basic scheme of switching, and by the provision of a suitable mechanical method of operation, it has been possible to provide a distinctly new type of dial switching. It is much too early to make predictions as to the extent of its ultimate use or the net improvements that will accrue from its employment, but it offers opportunity for shortening the switching time and for decreasing the maintenance. Apparatus has been manufactured, and the first trial installation in a dial central office is now going forward. Several specimens of pressure-testing plugs are being subjected by H. Baillard to life tests at elevated atmospheric temperatures A Power Amplifier Tube for Ultra -High Frequencies By A. L. SAMUEL Vacuum Tube Development THE development by the Laboratories of an amplifier tube capable of handling a moderate amount of power at frequencies as high as 30o megacycles per second now makes possible an appreciable extension of the usable portion of the radio-frequency spectrum. The use of conventional vacuum tubes at these very high frequencies has been found unsatisfactory because of certain effects which at lower frequencies are of secondary importance. For an appreciation of these effects, certain concepts are necessary. One of them has to do with the time required for the electrons to travel from the cathode to the anode within the tube structure. This time 344 is the so- called electron- transit time. At low frequencies it can be neglected; at high frequencies it must be considered. One effect it produces is a lag in the phase of the output current with respect to the grid potential. The calculation of this delay is complicated by an important distinction which must be drawn between the rate of arrival of electrons at the plate and the plate current. As an electron approaches the plate it induces in that plate an image charge. The magnitude of this charge varies with the proximity of the electron to the plate. The flow of current in the conductor to provide this charge actually constitutes the plate current. Viewed in this light the component of plate cur7u/y 1937 rent due to any given electron commences to flow when this electron leaves the cathode and ceases to flow at the instant of the electron's arrival at the plate. Nevertheless, the net effect of the transit time, as may be shown by a detailed analysis, is to produce an appreciable phase difference between the grid potential and the plate current. A further consequence of the finite transit time is that under operating conditions (that is with alternating potentials on the tube electrodes) the electrons arriving at the plate will usually have velocities greater than the velocity corresponding to the potential of the anode at the instant of their arrival. The excess energy corresponding to the greater velocity is obtained from the alternating component of the electrode potentials, and its dissipation at the plate in the form of heat decreases the useful output obtainable from the tube. Part of this energy comes from the grid circuit, and is responsible for the so-called input impedance or active grid loading. The practical effect of this input loading in an amplifier is to increase the power demands placed upon the input supply. Its effect is by no means negligible even at only moderately high frequencies, and at ultra -high frequencies this input loading becomes of major importance. A second important concept for the correct understanding of ultra -highfrequency tube design has to do with the increased importance played by the interelectrode capacitances and the lead inductances. The difficulties encountered in the use of the simple three -element tube as an amplifier at moderately high frequencies as a result of feedback or singing caused by the interelectrode capacitances are, of course, well known. Such difl'icul7uly 1937 ties are greatly increased at higher frequencies. They may be largely overcome by the use of a multi -element tube structure. At ultra -high frequencies, lead inductances common to both input and output circuits produce a similar effect, and so must be avoided in the tube design. The large charging current required by the interelectrode capacitances at high frequencies affects the cathode design. At low frequencies the rate at which electrons leave the cathode at any instant is identical with the rate at which they arrive at the anode. At high frequencies this is no longer true. The peak instantaneous emission may greatly exceed the value that would be required for operation under identical voltage conditions but at a lower frequency. The high charging current is also responsible for an increase in the resistance losses in the tube leads. The resistance of these leads is, of course, greatly increased at high frequencies because of the so-called "skin" effect. Resulting losses decrease the efficiency CONTROL GRID SCREEN GRID SUPPRESSOROR! Fig. sketch of the new high i- Perspective frequency double pentode - 345 of the tube and may, in a power tube, produce enough local heating to cause a more or less rapid deterioration of the lead-to -glass seals, which may ultimately destroy the vacuum. Short, tances, lead inductances, and electron transit time will be reduced in direct proportion to the reduction in size. Unfortunately, a reduction in dimensions without a corresponding reduction in all operating voltages is possible only at the expense of an increased demand on the emission capabilities of the cathode. The required emission per unit area must vary inversely as the square of the linear dimensions. Added to this is the increased demand caused by the high frequency charging currents already discussed. The available emission is fixed by the character of the cathode surface, and cannot easily be inFig. 2-One of the grid structures of the creased. A proportionate reduction in new tube shown slightly more than three cathode dimensions, therefore, is not times actual size feasible. Furthermore, the proportionate reduction of the anode dimenheavy leads are therefore required for sions would require an increase in the ultra- high- frequency operation. power dissipation per unit area The interelectrode capacitances to- again inversely proportionate to the gether with the lead inductances are change in linear dimensions. While the responsible for still another difficulty. heat -dissipating ability of the anode The frequency to which the input and can be increased in a number of ways, output circuits of an amplifier may be most of these will increase the tube tuned is set by the natural frequency capacitances. The high grid tempera formed by the interelectrode capaci- tures which may result from the retances and their associated lead in- duction in dimensions also makes ductances. For most practical pur- necessary the introduction of cooling poses the operating frequency of an provisions. Because of these effects amplifier must be well below these one must combine a reduction of divalues. This places an upper limit on mensions with the introduction of the permissible values that the inter- special mechanical arrangements to electrode capacitances and lead in- overcome the otherwise harmful efductances can have. fects of this reduction. Many of the factors which have All these factors have necessarily been discussed can be compensated by been taken into account in the dereduction in dimensions. It can be velopment of the new tube. As may shown, in fact, that if all the dimen- be seen in Figure i, and somewhat in sions of a vacuum tube are reduced in the photograph at the head of this the same proportion, the trans-con- article, it consists of two relatively ductance, plate current, and amplifi- large concentric metal cylinders and cation factor for fixed electrode po- two sets of tube elements diametritentials will remain unchanged while cally opposite each other outside the the values of interelectrode capaci- outer cylinder. The cylinders act as a - 346 7,1/y 1937 tional pentodes of similar ratings at much lower frequencies. Stable operation with some gain has been obtained at frequencies as high as Soo megacycles. When operating as a class A amplifier at i So megacycles, an output of i watt is obtained with the distortion 4o decibels below the fundamental. Under these conditions the stage gain is 20 decibels. Outputs of io watts with a plate efficiency of 6o to 7o per cent and a gain of io decibels are secured when this tube is used for class B operation. The development of this tube demonstrates that power amplifier tubes of the negative -grid type are usable at higher power levels and frequencies than have been reported previously. This type of development removes a practical barrier which, up to the present, has prevented the successful utilization of frequencies that extend above one hundred megacycles. 38 FREOUENCY = wU 150 MEGACYCLES z Fw 36 Eb=Es= N3 w> 34 SOOSOO VOLTS CC aDZ 32 z w? 30 rw 400 v ~ 28 LL w0F U' C w 26 =U gá 24 CC z á 320 22 2 20 60 Fig. 100 90 95 75 80 85 70 65 TOTAL SPACE CURRENT IN MILLIAMPERES 5- Input- output reaction determined experimentally at 15o megacycles www.americanradiohistory.com Rectifier for Telephone Power Supply By D. E. TRUCKSESS Equipment Development N small central offices the storage battery which supplies power for the telephone equipment is usually charged with a rectifier, such as the Tungar rectifier. Chargers of this type are equipped with two -element tubes and require a regulating device, usually including a rheostat, which is wasteful of power and frequently requires manual control to maintain the voltage within the limits required by the telephone circuits. It is often not practicable to provide manual control, particularly in outlying offices, and for such situations a device is required which does not demand continuous attention. This need has been effectively met by a new type of rectifier with grid -controlled tubes, which automatically maintains the charge in the battery, regulates its voltage and recharges it when necessary. The success of the device depends on the use of grid -controlled rectifier tubes, a recent development which has greatly extended the application of rectifier tubes to power problems. The addition of the grid does not give continuous control of the plate current as is the case in conventional three -element vacuum tubes but it makes it possible to control the time when the plate current begins to flow. This occurs when the grid voltage becomes less negative than the critical breakdown grid voltage of the tube. With grid voltages more negative than the critical value, plate current cannot flow. As soon as the current starts, 350 7111y 1937 www.americanradiohistory.com however, the grid loses control and the magnitude of the plate current is determined by the load impedance and plate voltage. In this case the current will continue to flow until the plate voltage is reduced to zero, hence an alternating current voltage applied to the plate lets the grid regain control every cycle. This makes it possible to control the output current of the tubes by changing the relative phase relations of the voltage applied to their grids and plates. When the grid voltage is in phase with the plate voltage the tubes will deliver maximum current by starting then at the beginning of each cycle, and when 18o degrees out of phase they will deliver no current since the plate voltage is zero when the tubes are to be started. The phase relation of the grid voltage can be changed by varying the resistance in one arm of a phase- shifting bridge circuit, which is the method used to regulate the rectifier's output voltage. The details of the RECTIFIER circuit are shown on Figure i. The output of the plate transformer TI is applied to the TI grid -control tubes vi AC and v2. Retardation SUPPLY coil LI is used in the rectifier circuit to filter the direct -current output. Automatic control is obtained by applying the battery b voltage to the screen T21 grid vacuum tube v4 plate -cathode resistance constitutes one arm of a bridge -type phase- shifting circuit, which includes two windings of the transformer T2 and the condenser ci. The voltage across the bridge is applied to the grids of the rectifying tubes through the transformer T3. As the battery voltage varies the regulator tube v4 varies the bias on the phase shifting tube v3, and changes its plate- cathode resistance. This shifts the phase of the grid voltage of tubes vi and v2, thus changing the output current from the rectifier in a direction to cause the battery voltage to return to the desired value, and in this way maintain a constant battery voltage. For manual control of the output the vacuum tubes v3 and v4 are disconnected and a rheostat is substituted for v3. If the rectifier output current exceeds its capacity the rectifier is converted to constant -current operation and the battery is charged at a constant rate until its voltage reaches the r LI --------R2 I TO VI STORAGE BATTERY V2 + LOAD T DRY CELL GRID BATTERY CI RI V4I111P7Or which amplifies the AMPLIFIER fractional volt changes L BRIDGE The volts. to several plate of this tube is Fig. 1-These rectifiers have two three -element grid-controlled connected to the grid mercury -vapor tubes whose output is automatically regulated of the three -element by changing the phase relations of the voltages applied to their grids and plates vacuum tube v3 whose 7uly 1937 351 desired overcharge value. This is accomplished automatically by the operation of a transfer relay, actuated by a current relay in series with the output of rectifier, which transfers the grid of tube v4 from resistor R1 to the series resistor R2, in the negative output lead. The regulator maintains a constant voltage drop over R2 and thus a constant current output. When the desired overcharge voltage is reached a high- voltage alarm relay releases the transfer relay and returns the circuit to constant-voltage operation. This permits the rectifier to be used in unattended offices as it can start automatically after a power failure, charge the battery and return to normal floating operation without any adjustments from an attendant. It also permits the rectifier to be used in other than telephone power plants where the load on the battery may vary from no load to several times the capacity of the rectifier. The headpiece shows a rectifier of this type which has a DC output of 8 amperes. It can be used at 132, 142, or 1.52 volts with only an adjustment of taps on the RI resistance. This regulating circuit will maintain the battery voltage constant within ± per cent for line voltage changes of ± io per cent, for load changes from no load to full capacity, and for room temperature changes from io to 40 degrees Centigrade. This is much better control than can be attained with unregulated rectifiers equipped with two-element tubes. / In this impact test for goggles, W. S. Hayford is observing the effect of dropping a steel ball upon the glass from any desired height in a metal tube 352 July 1937 shield between the input leads to the is made desirable by the ultra -highcontrol grid and the output side of frequency requirements which have the tube, as supports for the screen just been described. and suppressor grids and as a radio In spite of the unusual form of the frequency by -pass condenser between tube, electrically it is the equivalent them, and as low- impedance leads of two conventional negative -grid, interconnecting the two sets of screen pentode tubes. Its performance at freand suppressor grids. The control quencies as high as 30o megacycles is grids are of an unusual design, con- quite comparable with the performsisting of a cooling fin to which are at- ance of conventional tubes at much tached loops of tungsten wire encir- lower frequencies. cling the thoriated tungsten filament. Operating characteristics and conOne of these is shown in Figure 2. stants are listed in Table r. Special atThese control grids project through tention is directed to the values of the slots in the cylinders and are in interelectrode capacitances and lead turn surrounded by loops of wire at- inductances. It will be observed that tached to the inner and outer cylin- while the interelectrode capacitances ders and acting as the screen and sup- are low they have not been reduced in pressor grids respectively. The control grids are supported directly on ,/ 28 FREQUENCY = 150 MEGACYCLES their leads which project through one 26 E b =Es= face of the tube envelope. The semi 500 VOLTS cylindrical anodes are also supported N0 24 directly on their leads which project z 22 through the opposite face of the tube 0x envelope. This unusual construction LL 400 100 TYPICAL CONVENTIONAL TYPE TUBE 2 MIDGET RECEIVING TUBE I 80 N 60 = 50 3 O u. 0 40 12 80 N Z 30 0 320 ULTRA -HIGH FREQUENCY DOUBLE PENTODE (GRID -TO -GRID) 95 85 90 80 70 75 65 TOTAL SPACE CURRENT IN MILLIAMPERES 100 4- Variation in input resistance with Fig. operating conditions at I5o megacycles 20 z 5 4 40 50 60 80 100 200 500 300 FREQUENCY IN MEGACYCLES PER SECOND Fig. 3-Push-pull input shunting resistance as a function offrequency proportion in the reduction of operating wavelength. A more important feature, however, is the reduction of the lead inductances. For a tube which is to be used at ultra -high frequencies, certain characteristics not ordinarily considered are of particular significance. One of the most important of these is the active grid loss which, as already mentioned, comes about because of July 1937 347 www.americanradiohistory.com appreciable electron transit time. Figure 3 gives a plot of the push -pull input shunting resistance of this tube as a function of frequency. The value of 30,000 ohms at 15o megacycles is to be compared with ^-,000 ohms, a typical value for two conventional tubes in push -pull. At 300 megacycles the input resistance of the twin pentode is still above 6,000 ohms, while for conventional tubes it is so low as to make them entirely inoperative. The variation in the input resistance with the operating conditions of the tube for a constant frequency of 150 megacycles is shown in Figure 4. It is evident that if a high value of input resistance is to be realized, high anode potentials with low space currents must be used. The reduction in the filament-grid spacing made possible by the unusual construction is in a large measure responsible for the improvement in the input resistance just noted. A characteristic measurable only at the operating frequency is the interaction between the input and output circuits which results from the residual value of the grid plate capacitance. This TABLE I reaction differs from Operating Characteristics and Constants of the Double Pentode Tube that predicted for the Filament current (each side) low- frequency capacity 5.o amperes Filament potential (each side) 1.5 volts measurements on a Rated anode dissipation (each anode) watts 15 cold tube because of Rated screen dissipation (each side) watts 5 the inductance of the At Anode and Screen Potentials of 50o Volts and Anode Current screen -grid lead, and of o.03o Ampere -Characteristics of each Side because of the electron Transconductance 125o micromhos Anode resistance 200,000 ohms space charge. The reNormal control grid potential -45 volts action can be measInterelectrode Capacitances (When Properly Mounted) ured by observing the Direct control grid to control grid 0.02 micromicrofarad variation in the input Direct plate to plate Total control grid to ground (each side) Total plate to ground (each side) Control grid to plate (each side) o.o6 micromicrofarad 3.8 micromicrofarads 3.o micromicrofarads o.oi micromicrofarad from the tuning and loading of the output 0.07 microhenry circuit. Experimentally determined values Lead Inductances Total grid to grid Total plate to plate impedance resulting o o8 microhenry Rating as Class A Amplifier Maximum direct plate potential Soo volts Maximum direct screen potential soo volts Maximum continuous plate dissipation (each) 15 watts Maximum continuous screen dissipation (total) io watts Maximum output at iso megacycles with distortion down 40 decibels i watt Nominal stage gain at 15o megacycles 20 decibels Nominal control grid potential -45 volts Rating as Class B Amplifier Maximum direct plate potential 500 volts Maximum direct screen potential soo volts Maximum space current (total) iso milliamperes Maximum continuous plate dissipation (each) i 5 watts Maximum continuous screen dissipation (total) 10 watts Maximum output at tso megacycles io watts are given in Figure 5. The double pentode tube has been found useful as a high quality class A amplifier, as a class B amplifier, as a frequency multiplier, and as a modulator at frequencies of Soo megacycles per second and below. Its performance in these various modes of operation is quite comparable to the performance of convenJuly 1937 348 www.americanradiohistory.com I crly experimental crossbar switching equipment installed for development at the Laboratories. i "` >iié ail , -°, J, .">i%ji ,. rn -_ yj'"'+-Ii 1!lL11 1 s?it ',íqtrt;' II Effects of molecular strains in metals are studied with this \ -ray speclrometcr. III Testing the holding power of a drop -wire clamp. IV later crossbar switching equipment in a trial installation at the Troy central office, Brooklyn. The p wiewwiroSi+Smgy; t11f11 }i sl1lll ") ! ! .` 1311; fhd411 .I It114M 141 . 111,1: , 1:1 hl:Gl 11'1:1 114:14 , '1` 6a:I:1: 1í.1'E1: . 4 4444 C 1 r te Lt: 7 7 ':« â !` l r(1 rI f1F'/' r NSW t -1 7 r ir -4" rivA»ía%rr 41 ¡AI WWI& MAI ¡Wilt ¡Mir L tÌL ihm'1iits sa,_r-- - -= rp«=: :S`ñ: Mannaj -. 1 .i. .,e°rN1 I rtrr `74a www.americanradiohistory.com 11 A High -Quality Headset for Monitoring By F. S. WOLPERT Transmission Instruments Development N connection with electrical recording, transmission, and reproduction of speech and program material in the sound picture and broadcast fields, the ability to monitor in a satisfactory manner at various points is essential in the production of a high -quality performance. When recording sound for a talking motion picture, for example, it is essential to be able to determine at the time the recording is made as nearly as possible how the sound will be reproduced in the theatre when the picture is shown. In other words, it is desirable to see the action on the set and to hear the corresponding sound as it would come from the loudspeakers in 7u/y 1937 the theatre. For this purpose, a monitor room or portable monitor booth is frequently provided on the sound recording stage. There are many circumstances, however, such as outdoor locations, where the use of a monitor room or booth is impossible. It is also recognized that greater facility of recording operations may be obtained on the sound stage if the mixer can be seated in the open alongside the director rather than in the booth or distant monitor room. For these purposes, it is necessary to use a pair of headphones for monitoring in place of the usual loudspeakers. The work of a mixer consists in determining whether the sound picked 353 which have been avail- able for monitoring purposes have not been satisfactory for a quality job. Most tele- 1-The D-97689 moving-coil receiver as part of the D -97690 high -quality headset used for monitoring in the Fig. sound picture and broadcast fields phone receivers in common use give rather low response at low frequencies, a high peak in the neighborhood of I000 to 2000 cycles, and extremely low response at frequencies above these. Considerable prog- ress has been made within the past few years in the development, design, and manup by the microphone will make a ufacture of moving coil receivers for suitable sound record, adjusting the various applications. These receivers volume to a suitable amount, and are characterized by a wide frequency making sure that no undesirable range, uniform response, and very sounds are picked up. In addition to little non -linear distortion, so that his own personal judgment of sound they are particularly suitable for use quality he must be equipped with in monitoring sound records. A resound reproducing apparatus capable ceiver of this type, known as the of giving him a faithful indication of D-97689, has recently been developed the sound transmitted to the record- particularly for this purpose. It is ining machine. Very similar to the tended to replace the less efficient and sound recording job is the work of the more expensive moving coil receivers technician in operating speech input previously available. A photograph of equipment in broadcast studios or re- the new receiver forming part of the mote pickup points. In transmitting D -97690 high -quality headset is the program material over telephone shown in Figure i. lines, such as the extensive nationThe construction of the receiver is wide networks, it is important that shown in Figure 2. The moving coil the test room forces at the various element is made of aluminum ribbon intermediate offices be able to check wound on edge and insulated with a the quality of transmission in connec- varnish enamel which serves also as tion with identifying and locating an adhesive for holding the adjacent troubles and other service difficulties. turns together. It is cemented rigidly Up until recently it has been neces- to the diaphragm and the leads are sary, in order to monitor on a high- brought out between paper insulating quality basis, to make use of loud- washers. The coil is located in t