US1603805A - Filter circuits - Google Patents

Description

Oct. 19,1926. 1,603,805

H. RIEGGER FILTER CIRCUITS Filed August a, 1922 Patented Oct. 19, 1926.

UNITED STATES PATENT OFFICE,

HANS RIEGGER, OF IBERLIN-PANKOW, .GERM ANY, ASSIGNOR 'I'O SIEMENS .8: HELSKE,

AKTIENGESELLSOHAFT, OF SIEMENSSTADT,

CORPORATION.

NEAR BERLIN, GERMANY, A GERMAN FILTER cracurrs.

Application filed August 8, 1922, Serial No. 580, 69, and in Germany August 17, 1921.

It is known to use for. the transmission of electric oscillations a chain of filter circ u1ts which permit the passage of oscillations within a determined frequency range, but which suppress the oscillatory currents out*.

tive as well as capacitative coupling exists,

the self-induction and capacity at the points of coupling'being so selected that at the limits of the filter range the coupling becomes zero.

A chain of filter circuits of this type has the advantage that at the above-mentioned limits practically no oscillations are transmitted whereby the admittance of the chain beyond these limits is generally quite small while all frequencies located within these limits are well transmitted.

An embodiment of the present invention is shown by way of example on the accompanying drawing, in whic 2 Figure 1 is a diagrammatical illustration of the filter circuit chain, while Figure 2 shows the eficiency curve, and

Figure 3 shows a modification of the circuits shown in Figure 1.

The windings 1 at one end of the chain, being suitably excited transmit the oscillations through the chain of filter circuits to windings 2 at the other end of the, chain, by which latter windings they may be trans mitted to any desired receiving element. The

chain consists of two members 3 and 4:. Both members are of almost similar construction. The member 3, for instance, consists of two elementary oscillatory circuits of which one comprises the self-induction 5 and the capacity 6,while the other comprises the selfinduction 7 and the capacity 8. These two elementary circuits are coupled with each other inductively by the transformer windings 9 and 10 and capacitatively by the condenser 11,-and each of these elementary gircuits is tuned to a frequency to be transmitted to produce a well known two peak resonance curve. The windings 12 serve as a medium for coupling this chain member to the energy supply coil 1, and the windings 13 serve as a coupling medium with the second chain member 4. If the self-induction produced by 9 and 10 and the capacity 11 form for themselves a combination in resonance with the transmitted electric oscillations, the

resonance occurs at one of the two limits of the desired frequenc range, beyond which oscillatory currents s ould not be allowed to pass. In the second chain member 4, on'the other hand, the corresponding parts are so dimensioned that the resonance occurs at the other limit of the range. In this manner, an efficiency is produced as is illustrated by the curve shown in Figure 2.

The electromotive force in the supply windings 1 (Figure '1) may be maintained constant, the frequency of oscillations being however variable. If in. this case the frequencies are plotted as abscissae while the currents that are produced in the receiving wlndin s 2 are plotted as ordinates, the curve s own in Figure 2 will be produced. It is assumed in this case that the two systems are suitably coupled so as to slightly broaden the resonance curve within a very narrow band of frequencies, which is deslrablc in practice. For this reason the resonance curve shown in Figure 2 has been shown with well known two peaks closely located together, which represents for all practical purposes a resonance curve with a slightly broadened average peak having substantially the same maximum value between the frequencies (.0 and At the frequency w, resonance occurs at the coupling points a, b in chain member 3, and at the frequency (0 resonance occurs in the chain member 4 at the corresponding points. Consequently, no currents How in the windings 2 at the frequency w, or (0 Within the frequency the frequencies o, and w The resonance curve must also at these frequencies touch the abscissae line and it will, therefore, not considerably ascend between (0 and w, on the one hand and between 1, and w, on the other hand.

Such a modification is shown inFigure 3. Aside from the coupled members '3 and 4, such as are shown in Figure 1 the chain further contains the coupled members 23 and 24. The con lings between the points a and b in the mem er 3 becomes substantially zero value at the vfrequency w, (Fi' re 2),

between the points a and d in mcm er 4 it becomes zero at the frequency (0 both similar to the manner explained with reference 1 each member, consisting of a plurality of elementary oscillating circuits electromagtnetically and electrostatically coupled with each other, the self-induction and the capacity at the coupling points being suitably selected to reduce the degree of couplingat a predetermined frequency to substantially zero value.

2; An improved chain of electric filter circuits, comprising a plurality of members each member consisting of a plurality of elementary oscillating circuits electromagnetically and electrostatically coupled with each other, the self-induction andthe capacity at the coupling oints being suitably selected to reduce the egrce of coupling at a predetermined frequency to substantially zero value, one member of the chain having said zero value at the lower limit, and another member having said zero value at the upper limit of a predetermined range of frequencies.

In testimony whereof I afiix my *si ature.

HANS RIEG ER.

Images