US2505024A - Wave translating circuits - Google Patents

Description

April 25, 1950 M. F. WlNTLE' WAVE TRANSLATING CIRCUITS Filed May 20, 1947 300 YOLZS INVENTOR MA LCOL M F- W/NTLE ATTORNEY Patented Apr. 25, 1950 Malcolm Fran'k 'aWintle, London, England, :assignal rtol'lnternational Standardi'Electric :Gor-

e1oiation,;:New York, N

Belaware Y., a corporationtoi Application May'20, 1947, Serial No; 749327 In Great Britain J-u'l-y fi, 1946 sciaims. (01.25.0527) l The present invention-relates to electric .wave translating circuits employing thermionic Vvalves.

The principal :object of the invention is to pro- .videa thermionic valve' circnit which can-tbewused for-aperiodic frequency doubling without the necessity for employing :anyirequencyedependent auxiliary devices {such as tuned circuits oryfilteizs. The (circuit can therefore-be used for frequency doubling 'over a wide range of frequencies with outsany modification.

Thisiob ject is achieved according to the inven tion .by providing an. electric wave translating circuit comprising a thermionic .valve having an anode-a cathode, avcontrol. electrode, an ,anode resistancefia cathode negative feedback resistance, means for applying a signal voltagcuto the said control electrode, andlmeans for deriving .an output voltage from the. said anode, in which the said resistances. are of such values that the curve of instantaneous ano'devoltage versus instantane-- ous control electrode .voltagehas a substantially constant negative slope overthe lower half o the working range, and a substantially constant positive slope over the remainder of the working range. v As Willbe pointed out laten'the circuit arrangement according to the invention has-other useful applications 'hesides'rrequency doubling.

The invention will be 'described with reference to' the accompanying drawing in 'which--'- Fig. 1 shows a simplifiedbasic schematic of'a thermionic valve circuit used explain the principles of the invention;

Pig 2shows characteristic curves obtained-with the circuit ofFig. 1.;

'Fig. 3 shows oneexainple of a practical circuit according to the invention; and

Fig-.4 shows two doubling stages :in tandem illustrating two methods f coupling :such stages together.

In the experimentalzcircuit "shown in Fig.1, 'a valve TI has its-anode connected "to the positive hightensionterminal 2 through a high resistance 3 of value R1, and its eathodeto-ahe negative high. tension terminal 4 through *a'hi-gh resistance ilof valve R2. -Let Vg' be the steadyrvoltage rain plied :between theqcontrol gri and terminal-A, andlet Vahe the correspondin altagemeasured hetweenztheanodeand'terminal 4. V 5;

v The resistances Br and shouldqboth be large compared with the reciprocal of 1 thexmutual {09.nductance of the valve. have values between ,.0.-1 and .Lo megohm.

. When applied thevoltage-lg .sislzero he anode .cnrrent :will be limited toa small yaliielqy there;

Theyrmis t,i rexamp V I2 sis-tances R1 and R2, an. the valve -w;il-l be biased nearly .to the cilt-ofia,pcint by thepotentialzdrop of theanode:currentyiiowing in R2, and cinder this condition the control. grid is negative to. thecathode, and there is'nogrid cuiment.

As the applied voltage V increasespositively from. zere, 1a-large1:negative teedback is. provided by the resistance Rzin series v 'iththe-cathode, and the negative potentialdifierence doe-tween the control gridand the "cathode decreases :at amuch slower rate,-en aocountofthe negative-feedback.

age .at sthe: :same time falls continuously owing; [the increase in the anod-ecurrent. When the contraband-cathode {cotentia-l difference reae-hes zero, the potential difference between thesanode and cathode will :be very low, possibly-a few volts, because Ri-ran-d ,-Rz will both be-l'arge compared with the-anede-oathode'impedance of the valve in this -condition ;so that they will absorb the major. proportion of the-applied highstension voltage.

At or iustbefore the potential difierence between the control-grid and the cathode has reached nerd-grid current :will 'begin to -flow. When this occurs the controlr-grid-cathode impedance becomes comparable w th -.or smaller than R2 so, that practically-the wholeeof any further signal lvoltageincreases new .occur across R2. There must therefore be corresponding reductions in the voltage drop across R1, so-that the anode-voltage now rises-again as the signal voltage is increased .Fig. 2 gives characteristic purves showing the relation foundqhetween'll iandvsdn the arrangement of Fig. =1, in which the valve employed is the, 1 3F150 type pen-tode; with the :screen grid and suppressorigrid connected to theanode. The resistances Brandi-B52 both had the same value 370,000sohms. The curves 5A,- B and .6 represent the characteristics obtained when the =high ten sion: operating voltage had. :the :values i=0!) --vo1ts, 200 volts and 300 volts, respectively. [All the ourves areisim-ilar in shape and comprise two practically straight inclined portions-with opposite slopes, connectedby a comparatively a5 nei hbcnrhoodgdfithe munved; portion. ".Thus'n the control grid be biassed so that the anode voltage has the minimum value M, and a small alternating voltage e sin wt be applied to the control grid, the anode voltage will be given by ke sin wt= ke (10OS Zwt), where k; is a constant. Thus a double frequency component will be obtained from the anode, accompanied by a direct current component ke which can be removed by a blocking condenser, or by a transformer. The action of the circuit is independent of the value of m. If e is not small, the double frequency component will still be obtained but it will be accompanied by other multiple frequency components of small amplitude.

It will be noted that since the slopes of the characteristic curve are practically 45, the bias value corresponding to the minimum point M will be nearly equal to E/2, where E is the voltage of the high tension source. In practice the proper bias voltage is slightly less than E/2.

1 If with this same bias. a train of short unidirectional voltage pulses be applied to the control grid, it will be seen that corresponding output pulses will be obtained which will always be positive pulses, whether the input pulses are positive or negative. Moreover, if the amplitude of the input pulses is large, the output pulses will be of nearly the same amplitude. Thus if a train of mixed positive and negative pulses be applied, all the negative pulses will be inverted, and the output train will consist of all positive pulses.

'If the bias be adjusted so that the working point comes on the negative sloping portion of the curve, then the valve operates as an ordinary amplifier with a gain ratio of substantially unity, and a phase change of 180. If the bias is adjusted so that the working point comes on the positive slope, the same results will be obtained except that there will be no phase change. Hence by providing simple means for changing the value of the bias voltage, the phase of a signal wave may be reversed when desired. This facility, and that mentioned above by which positive output pulses can be obtained from pulses of either sign will be useful, for example, in synchronising time bases for cathode ray tubes.

Referring again to Fig. 2, it will be evident that since the anode voltage is always equal to the applied high tension voltage E less the potential drop produced by the anode current in the resistance R1, the curves relating the anode current to the applied voltage V will be similar to the curves A, B, C, but inverted, so that the point corresponding to M will be a maximum instead of a minimum. It follows that if a relay (not shown) be connected to series with R1, it could be designed or adjusted to operate only for currents exceeding a given value, and then the relay could only be operated by the application of input voltages Vg within a certain relatively narrow range, centered about the value corresponding to the point M.

Fig. 3 shows a practical frequency doubling circuit arrangement according to the invention which includes all the elements shown in Fig. l. The control grid of the valve is connected through the secondary winding of a transformer 6 to the junction point of two resistances and 801 valves R3 and R; connected in series between terminals 4 and 5. The high resistances R1 and R2 should be equal. The waves whose frequency is to be doubled are applied to input terminals 9; and I connected to the primary winding of the transformer 6. The output double frequency waves are obtained from a terminal I I connected through a blocking condenser I2 to the anode of the valve I.

The resistances R3 and R4 provide the bias for the control grid of the valve, and should be sufiiciently small compared with R1 and R2 so that thebias potential is not appreciably affected by the grid current. If E is the potential of the high tension source, then the bias produced by R3 and R4 is R4.E/(R3+R4) and should be adjusted to the value corresponding to the minimum point M of the curve shown in Fig. 2, which applies to the value of E used. As already explained this value will be slightly less than E/2. Then in accordance with the explanations already given, the desired double frequency waves free from the direct current component will be obtained from terminal I I'.

The secondary winding of the transformer 6 should present a low impedance to the grid current for reasons already explained.

It'will be evident also that if trains of short unidirectional pulses are applied to the control grid of the valve l through the transformer 6 (or in any other conventional way), the output pulses obtained at terminal II will always be positive pulses, irrespective of whether the input pulses are positive or negative.

Furthermore, by making the resistances R3 and R4 adjustable, the operating point may be placed on the straight part of the Fig. 2 characteristic either on the positive or on the negative slope, as desired, and then normal amplification will be obtained without appreciable distortion.

A phase change of will be obtained if the negative slope is used, but none with the positive slope.

Fig. 4 shows two stages of frequency doubling and illustrates two other methods of coupling the doubling circuit to a proceeding stage.

The first doubling stage comprises three elements IA, 3A and 5A similar respectively to elements I. 3 and 5 of Fig. 3. The input waves are applied through a cathode .follower valve I3 having its anode connected directly to the terminal 2 and its cathode connected to terminal 4 through a resistance I l. The control grid of the valve IA is connected directly to the cathode of the valve I3. The control grid of this valve is connected to ground through a high resistance I5, and the input waves are applied to terminals I6 and H, the former of which is connected to the control grid, the latter being connected to ground. This cathode follower stage provides in well known manner the desired low impedance input circuit for the doubling valve IA. The resistance I4 should preferably be chosen so that the desired bias of approximately E/2 is obtained on the control grid of the valve IA, or suitable auxiliary biassing arrangements (not shown) should be provided in any known way. The input waves could, for example, be derived from thevalve I of Fig. 3 by connecting terminal II to terminal I6.

The second doubling stage of Fig. 4 comprises a set of elements IE, 33, 53, H3 and IZB similar to the same numbered elements of Fig. 3. The control grid of the valve 513 is connected directly to the anod of the valve IA, thus obtaining' the proper bias automatically, since as already explained, the anode potential at the minimum point M is approximately equal to E/2.

It will be evident that any number of frequency doubling stages could be coupled together and to the source of signals by any of the coupling arrangements which have been described with reference to Figs. 3 and 4.

It may be added that if the input; signal voltage is large, so that some distortion is unavoidable as a result of frequency doubling, it is preferable to use large values of R1 and R2, since there will then be less grid current and a smaller loading of the input circuit. The sloping portions of the curves of Fig. 2 are then straighter, since for the negative slope the negative feedback intrcduced by R2 is greater, while for the positive slope the voltage across the valve, which varies in a non-linear fashion will be a small proportion only of the applied high tension voltage.

When, however, the input signal voltage is small, and it is desired to avoid distortion as a result of doubling, the range of the parabolic portions of the curves of Fig. 2 may be extended by making R1 and R2 rather smaller. This provides less negative feedback and increases the effectiveness of the curvature of the grid-current characteristic of the valve.

The advantages of the circuit according to the invention may be briefly listed below:

1. Since there are no filters or tuned circuits the circuit will operate as a frequency doubler over a wide range of frequencies.

2. If the input amplitude is small, and if the bias of th doubling valve is correctly chosen, the output will consist substantially only of the double frequency waves without other multiple frequencies of appreciable amplitude.

3. The circuit will produce a train of positive output pulses in response to a train of positive or negative input pulses, and if the input pulses are of large amplitude, the output pulses will have substantially the same amplitude. It is not necessary that the adjustment of the bias should be very accurate for this application of the circuit.

4. The circuit may be used as an ordinary amplifier with a gain ratio of approximately unity, and by a simple change of grid bias the phase of the output waves may be changed by 180 with out appreciably afiecting the amplitude.

What is claimed is:

1. An electric wave translating circuit for frequency doubling comprising a thermionic valve having an anode, a cathode, a control electrode, an anode resistance, a cathode negative feedback resistance, means for applying a signal voltage to the said control electrode, and means for deriving an output voltage from the said anode, in said anode and feedback resistances being of the same value and large compared to the interelectrode resistance between the cathode and grid such that the curve of instantaneous anode voltage versus instantaneous control electrode voltage has a substantially constant negative slope over the lower half of the working range, and a substantially constant positive slope over the re mainder of the working range.

2. A frequency doubling circuit according to claim 1 comprising means for biassing the said control electrode close to the potential of said cathode such that when the applied signal voltage is zero the anode voltage is a minimum.

3. A circuit according to claim 1 in which the signal voltage is derived from a low impedance source.

4. A circuit according to claim 1 in which the signal voltage is applied to the control electrode through a transformer having a low impedance secondary winding.

5. A circuit according to claim 1 in which the signal voltage is applied to the control electrode through a cathode follower amplifying stage.

MALCOLM FRANK WINTLE.

REFERENCES CITED The following references are of record in the file of this patent:

UNITED STATES PATENTS 40 Number Name Date 2,230,243 Haficke Feb. 4, 1941 2,256,085 Goodale Sept. 16, 1941 2,408,261 Kakatos Sept. 24, 1946

Images