US2594535A - Multiple channel electronic switch - Google Patents

Description

A ril 29, 1952 2,594,535

$- BERTRAM MULTIPLE CHANNEL ELECTRONIC SWITCH Filed April 26, 1944 2 SHEETS-SHEET 1 20" SWITCH 7 [4 H INPUT 41 SWITCH OUTPUT /6\ l3 SWITCH/NC /2 22 SIGNAL INPUT ill m 0* NWfi l5 F/ a 1 4 I 0 BIAS SECTION I Es W T 5 2 TUBE l2 CUT-OFF EC CUT-OFF SECTION 1 OFF TIME +L}fi 0A/ TIME CUT-OFFSECT/ON TCHl/VG T/NE TUBE l2 CUT-OFF 0 BIAS SECT/ON 2 S/D/VEY BERT/MM INVENTOR ATTORNEY A ril 29, 1952 s. BERTRAM 2,594,535

MULTIPLE CHANNEL ELECTRONIC SWITCH Filed April 26, 1944 2 sHEETs-s1-1EET 2 I Patented Apr. 29, 1952 UNITED STATES PATENT OFFICE I MULTIPLE CHANNEL ELECTRONIC SWITCH Sidney Bertram, Los Angeles, Calif., assignor to the United States of America as represented by the Secretary of the Navy Application April 26, 1944, Serial No. 532,915

1 Claim. 1

This invention relates to a multi-channel electronic switch.

In the past, many methods of switching multichannel devices have been suggested but they have,-in general, been rather complicated and inflexible. When it is desired to scan the outputs of a plurality of channels (or conversely, distribute the output of a signal channel) at relatively high frequencies, my invention provides an excellent means for doing so.

One of the outstanding advantages offered by my invention is the manner and ease with which the switching sequence is obtained. It differs from many prior switches in that no oscillating circuit is built into it, the switching signal being obtained from an external sinusoidal source. A phase shifting network is utilized to produce the switching sequence so that each switch element receives a switching signal displaced in phase by an amount corresponding to its position in the switching cycle.

Another advantage of my invention is that the various switch elements are inherently independent of one another; with interaction between switch elements made negligible,

A further advantage of my invention is that the basic switch element is very simple and contains but two tubes.

Still another advantage and object of my invention is that the on time of any switch point can be varied, if desired, to separate adjacent points orto cause them to overlap.

Yet another advantage and. object of my invention is an electronic switch controlled by a sinusoidal input, the versatility of which makes it available for a wide variety of purposes.

A still further object is an electronic switch composed of a simple basic element to which additional identical elements may be easily added for any desired number of channels.

A yet further object of my invention'is a multiple electronic switch containing no thyratron tubes and which is relatively free from noise.

And yet another object of my invention is an electronic switch which may be utilized for a variety of purposes, including distribution, wave analysis, monitoring, multiplex signalling, and secret communications, etc.

Another object is to provide an electronic switching system in which the rate of switching from one element to another in the system may be varied by the simple expedient of varying the frequency of the applied control source.

Still another object is to provide an electronic switching system in which the "011 time of each element in the system is controlled with exceedingly great exactitude.

A more specific object is 'to provide an electronic switching system in which all switch channels are controlled from a single sinusoidal source of voltage, such voltage being supplied to each of the various channels through a phase shifting network whereby the steep portion of the voltage wave can be used to very accurately control the on time of all switch channels in the system.

In the drawings:

Fig. 1 is a schematic diagram of the basic switch element.

Fig. 2 is a diagram graphically illustrating the operation of the basic switch element.

Fig. 3 is a schematic diagram of a, ten-channel switch, with seven channels omitted, and

Fig. 4 is a diagram showing the phase shifts introduced in the control voltage source by the type of phase shift network used with the ten channel switch shown in Fig. 3.

The basic switch element, shown in Fig. l, is.

designed for use on a single channel and the complete switch is constructed with a plurality of elements identical with it together with an appropriate phase-shifting network. It is comprised of two tubes II, I2. Tube I I is a high mu duplex triode and I have found that a GSN'I performs satisfactorily. Tube I2 is a pentode voltage amplifler tube of the sharp cutofi type, such as a 6SJ'7. The plates of tube I l have a common connection to the control grid of tube I2 and a resistor I3 is connected commonly into the plate circuit of tube II and the grid circuit of tube I2.

The two grids of tube I I are connected through high resistors I4 and I5 to the switching signal input terminals I6. Static bias for the grids of tube II is obtained from a rheostat I! at the lower end of voltage divider I3. This connects to the grids of tube II through center tapped resistor I9 which provides an electrical center for the signal input. Of course, if the signal input is balanced about an available central point, resistor I9 is unnecessary.

The switch input, supplied at terminals 20, is applied to the grid of tube I2 through a resistor 2I. This should be a high resistance and while this decreases the sensitivity, it is necessary to prevent a low impedance device feeding the circuit from short circuiting the switching pulse. A plate load resistor 22 is included in the plate circuit of tube I2 and the individual switch output is taken from this resistor at terminal 23. The suppressor-grid of tube I2 is tied to its 3 cathode and screen bias is supplied by a variable connection 24 to voltage divider id.

The operation of the basic switch element is exceedingly simple. Tube it is biased beyond cutoff when no switching signal is applied, and thus since tube i! has no efiect upon it, tube l2 operates like a normal amplifier. If, as is the case, a sinusoidal switching signal is supplied at terminals I6, both sections of tube ii are cut off during the portion of the cycle in the neighborhood of zero. However, there will be considerable portions of the cycle during which the sections of tube it become conducting, one section conducting during positive swings of the cycle and one during negative swings. In either case, when this occurs, the plate current of tube ll flowing through resistor 93 causes the bias on tube 52 to increase beyond the cutoff value.

This operation is illustrated diagrammatically on Fig. 2. The two sections of tube l designated section 1 and section 2. where:

Ec=the static grid bias on tube i l Ec =the cutoff voltage on tube i i E =the grid voltage on tube ll when tube 52 is effectively cut off.

Es t-he peak amplitude of the switchin signal.

As has been stated, tube i2 operates normally when both sections of tube ii are biased beyond cutoff. For instantaneous valuesof signal voltage (es) greater than the difference between EC and Ec (the static bias and cutoii bias of tube II), but less than the difference between EC and E60 (the static bias and cutoff bias of tube t2), the bias of tube E2 is changing but it is not cut off. Therefore, this region of the cycle represents the switching time. This time may obviously be varied as desired by changing the peak value-cf the switching signal ES (thus changing its slope). When the instantaneous value (es) of the switching voltage. increases beyond the difference between EC and E62 tube 52 is cut oil and remains so until the switching signal again decreases below this value. The same sequence of.events occurs on the negative portion of the switching cycle with respect to section 2 of tube H. It should be noted in this connection, that the steep portion of the sinusoidal curve is being utilized which obviously gives a very accurate control as to the time. at which the tube is switched on and off.

In connection with the switching time it should be noted that resistor l3 should be small enough so that residual tube currents beyond effective cutoff of tube ll do not detrimentally afiect the operation of tube l2. Additionally, the plate voltage of tube ii should be kept low to increase the efieotive mutual. conductance of the tube for low plate currents.

Since the on and off time or" the basic switch element is controlled by the phase of the sinusoidal switching signal, this samesignal can be utilized to operate a plurality of identical elements or channels in sequence, providing a phase shift is introduced into that signal before it is applied to each next adjacent switch element.

.Aphirality, namely, ten of such channels are illustrated in Fig. 3, although it is to be understood that any desired number could be utilized.

In Fig. 3, only the first two and last of the ten channels in the switch are illustrated, the diagram being broken between the second and. last channels. However, the remaining seven channels are the same as theones shown. Tubes 25a and 25b, in the first two channels and tube 4 257 in the tenth channel correspond to tube H in Fig. l and similarly tubes 26a, 26b and 257' correspond to tube l2. Although only one switch tube such as Zta operates from switching tube 25a, it is obvious that a plurality of such switch tubes could be so operated if desired.

A plate load resistor 21 is made common to all ten elements and the cathodes of tubes 25a-257' are tied together.

A sinusoidal source of potential which functions as a switching signal is supplied at terminals 2:3,and feeds the primary of transformer 29. One set of taps on the secondaryof transformer-29 is connected to rectifier to supply direct current to the screen and plate of tubes 2 341-267 in their respective channels.

A filter composed of choke 3i and capacitor 32 serves to eliminate hum from the direct current supply.

Voltage divider 33 corresponds in function to divider E8 in Fig. 1. Screen bias to tubes 26a-2Gy' is supplied by variable connections 3la3dy' oil divider 33.

Resistors 35a-35j correspond to resistor 43 in Fig. l, resistors 3566-15 correspond to resistor 2| and switch input terminals 31a-9' correspond to terminals 20.

Another set of taps on the secondary of transformer 25 feeds over conductors 38, 39 to a phase shifting network having two parts 46 and 3E.

' The output of all of the channels is supplied to the grid of a cathode follower tube 4|. Thus the outputs of all ten channels can be taken across cathode resistor 42 of this tube and supplied to the multi-switch output terminals .43.

In order that the proper phase shift be introduced into the switching signal, so that each element operates in its proper sequence during the switching cycle, the phase shifting network ail-49 is used to supply the properly phased voltages to the respective grid inputs of tubes 25a-257'.

The basic element of each of the network parts 40, 40' is a 11' section composed of series resistors 44 and capacitor :35. This section is designed to give the desired phase shift (/N, where N is the number of basic elements or channels in the switch), which in the illustrated case is 18, when properly terminated in its characteristic impedance. These sections are connected together and it isseen that individual capacitors 45 operate in adjacent 1r. sections. Since there is a large attenuation inherent in such networks, it is formed into two parts in order to get the total 180 coverage. This requiresv an input phase shifting arrangement consisting of capacitors t6 and resistors 47 to initially shift the phase of network 49' ahead of that of the source at terminals 28 by 45, and capacitor 43 and resistors 49 to shift the phase of the other network as behind 45. Thus a 99 shift between corresponding points on the two network parts is obtained.

Network parts 48, 40' are balanced and a center tap 58 from each section is led back (but not shown on the schematic) to the center tap on the secondary of transformer 28.

The voltages to the grid terminals of the switching tubes 25a-257' are taken respectively from adjacent sections of the network parts 40, Ellat terminals Sic-Sly through high, grid current limiting resistors 52.

For all except the lastsection of each network part, bleeder resistors 53 are used to equalize the signal voltage to tubes 25a-257', thus compensating for the attenuation inherent in the phasing network. Thus resistors 52, 53. perform the function of resistors l4, l and 19 described in connection with the basic switch element :shown in Fig. Resistors 54 serve as the resistance parts of the. terminating impedances for thenetwork parts 4.0, 40. Capacitors 55 combine the functions of shunt impedances for the vIlast section of each of the network parts 40, 40'

and the reactive elements of the terminating impedances.

1 tion could be combined into a single resistor and the other center taps 50 in each section eliminated.

In the operationof the embodiment shown in Fig. 3, the sinusoidal switching voltage wave from the source applied to terminals 28 and transformer 29 is impressed across the phase shifting network parts 40 and 4B. The individual sections of these networks apply this. voltage to the grids of tubes .25a- 257' but with different degrees of phase shift as previously explained. That is to say, and by reference to Fig. 4, wave A in phase with the voltage source is applied to the grid of tube a; wave B shifted 18 applied to tube 25b; wave C shifted 36 applied to tube 25c; .wave D shifted 54 applied to tube 25d; wave .13 shifted 72.applied to tube 25c and so on with progressively increasing phase shift through all ten channels, tube 257' in the tenth channel receiving a grid input voltage J 162 out of phase with the voltage applied to the first section of the phase shifting network part 40.

Tubes 26a-"-26y' will therefore be switched on and ofi and in succession for periods a to 7' respectively as seen from Fig. 4, twice during each cycle of the applied voltage wave as explained in connection with Fig. 1. As each of tubes 26a-26 are switched on in succession, any input signal appearing at terminals Sic-317 of these tubes will be amplified in the tubes, and then applied in succession to the grid of tube 4|. These signals then appear, in order, across output terminals 43.

In connection with the apparatus shown in Fig. 3, the speed of switching can be varied by changing the frequency of the A. C. source by any well known means.

The invention disclosed herein may be utilized in many ways. One use for which it has proven very successful is as an analyzer for a soundfrequency spectroscope. In this case, it is connected to scan the output of a plurality of fixed frequency analyzers to present the information thus obtained on the screen of a cathode-ray tube. For this use the linear sweep of the cathode-ray tube is synchronized at double the frequency of the switching signal. The control voltage for such purpose is obtained from the plates of one of the tubes 25a-257 in the several channels or, in the case of line frequency switching, from the rectifier tube.

It may also be used as an analyzer or distributor, in which case the inputs of the switch elements are connected to a common channel whose output is to be distributed. The outputs of the various switch elements are applied to separate channels. Thus, the input signal will be distributed amongst the output channels in a periodic time sequence.

To adopt the switch for monitoring a set of circuits, as for example a ten meter system, a resistor would be placed in series with each meter such that the voltage drops would normally all be equal. These ten voltages would be scanned by the switch and placed on a cathode-ray screen as is done in the sound spectroscope. The horizontal sweep would be divided into ten sections with one section being allotted to each metering circuit. Deviation from thenormal in any circuit would be evident on the cathode-rayscreen and could be interpreted by means of a' calibration. I

Additionally, the switch is ideally suited for use in multiplex signaling. It would be used both as a selector and a distributor, and because the basic switch elements are independent, the individual channels could be assigned and located at widely scattered stations. Each station would have a distributor-type switch element placed on the channel assigned to it and the output of this switch might 'be made to actuate a speaker, relay or other receiving device. A selector-type switch element with a mechanical switch would be provided for each station so that the channel might be changed at will. Thus, if an operator at one station desired to signal another station, it could be done by connecting his selector-type switch to the channel assigned to'the other station. In this way, an input at the operators station would produce a response at the other station without affecting the stations assigned to the remaining channels. g

The switch also is adapted for use in an audiofrequency communication system. In this case, the switching frequency would be made quite high with respect to the highest audio-frequency used, which fact would require the use of a wide-band communication channel. However, it would make possible the transmission of several messages on the same channel.

In conclusion, it is to be understood that various changes may be made in the illustrated embodiments of the invention without departing from the spirit and scope thereof.

For example, an alternative form of the basic switch element would utilize a tube with two or more controllable grids, such as a 6SA7, in place of tube I2. In this case, the switching signal would be applied to one control grid, while the switch input would be supplied to the second control grid. In the specific case of a GSA? tube, grids labeled l and 3, respectively, have been found very suitable for these purposes. This alternative form has the advantage of completely isolating the switch input from the switching signal.

Although the phase-shifting network described above produces excellent results, there may be cases in which slightly different system would prove more satisfactory. One such system includes taking the switching signal for the individual elements from the secondaries of separate transformers whose primaries contain R-C networks designed to shift the phase without changing the amplitude. Chokes might also be used in place of the transformers, but such a system would require more resistors and capacitors to provide a balance point.

Another phase-shifting system which can be utilized is an iterative network with series of capacitors and shunt resistors.

Another alternative system of phase-shifting would include the use of an R-C circuit, with or without a reactance tube, to provide a phase shift equal to the maximum required by the switch. Intermediate phase-shifts could then be obtained by combination .of. the-.hrigin'al with the .shiited signal.

Additionally, a systemmightutilize a reactaxioe tube for obtainingan essentially quadratureivoltage to be combined with theoriginal ,tozobtainla voltage of any intermediate *phasexshift. t-This'gis one which could be.made.relatively-independent of the switching frequency.

' Having thus: fully describedamy invention;- I claim:

A multiple channel electronic switch comprising a plurality of.pairs of switching. tubes, eachof said switching tubes. containingat least ananode, acathode, and a control.grid,;means.connecting the cathodes. of the tubes on each. said pairsE'of switching.v tubes together and theanodes' ofxthe .tubes in :each ofsaid pairs of-switching tubes togetheiybiasing: means for rendering said switching tubesnormally non-conducting, a switchtube for each 1 of. said pairs 'of switching tubes, each of saidswitch tubes having. at least an anode; a cathode, and: a control grid, means connecting the control grids of ..said switch tubes to the anodes ofv said pairs of switching tubeslrespectively, a plurality of resistances each havingone end. connected to the anodes of one-of said pairs of switching tubes and the othenend connected to the cathode of said associatedsw-itch tube, a plate load resistance for each switch tube having one end thereof connected to the anode ofsaid switch tube; first and secondsources of anode power connected to.1.said ;.plate .-load-: .rresistances and-to .the cathode, of :saidiswitching; .tllbBSlrIB- 'spectively, whereby. conductiorrroizsaid switching tubes renders said switchztubes;noneconducting, a. source. of. alternating; current; phaseeshifting means i connentedzito said. source iofjzalternating current producing a plurality 'of'. individualized alternating current voltages: .having incremental phase; :displacements: with .respect. to .:each other and magnitudes in excess of that ofsaidzbiasing -means, andmeans for impressing: said'individualized alternating. currentvoltages-;on-the; grids of "said: plurality :of pairs 'of-..'switching tubes respectively, wherebysaid switching tubes are. sequentially rendered .non-conductingzzwhen the alternating current voltage impressed on the gridsicf .each oftsaid pairs of switchmubesu-is.near zero.

. SIDNEY-l-BERTRAM.

REFERENCES; CITED The following references-are-of recordr-in-ithe file of-this patent:

UNITED. STATES PATENTS Number Name cDate 2,048,081 Riggs -7 July: 2,1,:1936 2172;354 uBlumlein Sept. :12; 1939 1 2,213,941 T. Peterson f Sept; 3,194!) 2,262,838 :Deloraineetal Nov: :18, :1940 2,363,062 Hartley 7.. "'Nov.-;21,.1-944 2,400,822 Hansell t'May 21331946

Images