US2978628A - Constant d.-c. voltage output circuit - Google Patents

Description

April 4, 1961 E. .1. DIEBOLD 2,978,628

CONSTANT D.-C. VOLTAGE OUTPUT CIRCUIT Filed July 27, 1955 2 Sheets-Sheet 1 IN V EN TgR. Eaumea Jay/v we; a

April 4, 1961 E. J. DIEBOLD 2,978,628

CONSTANT D.C. VOLTAGE OUTPUT CIRCUIT Filed July 27, 1955 2 Sheets-Sheet 2 INVENTOR. DW/9/60 day/v DAL-BOLD BY /1M,% M XW United States Patent CONSTANT D.-C. VOLTAGE OUTPUT CIRCUIT Edward John Dieboltl, Ardmore, Pa., assignor t0 LT-E Circuit Breaker Company, Philadelphia, Pa., :1 corpor. tlon of Pennsylvania Filed. July 27, 1955, Ser. No. 524,778

2 Claims. (Cl. 321-16) her of volt seconds for any voltage greater than a predetermined magnitude.

This application is a continuation in part of my copending application, Serial No. 312,053, filed September 29, 1952, and now abandoned.

The principle of my invention is to apply the input of an alternating voltage source of variable voltage magnitude to a winding of a saturable type reactor. The saturable type reactor is then so designed as to be too small to be completely unsaturated for the lowest voltage which the alternating current source will ever reach. Therefore, a voltage having a pulse shape, since the reactor will be saturated for a portion of the cycle, will appear on the winding of the saturable reactor which has this alternating voltage applied to it.

As is well known, the volt second area contained in this pulse will have a constant value. That is to say, that if the voltage is of a low magnitude the pulse will be relatively wide and of small magnitude. Conversely, if the alternating voltage is of a high magnitude, the pulse will be of a relatively short duration and will be of a high magnitude. However, it is essential to realize that the area under either of the pulses will be constant. If now this pulse is rectified, in view of the constancy of the area contained in the pulse, it is clear that the average voltage value of the pulse will be constant and will be determined by the design of the saturable reactor which determines the area under the pulse.

This fact becomes apparent when one considers that the average value of the voltage will be equal to the integral of the voltage times the differential of the time, the integral being taken over limits defined by the pulse width. Since this integral merely defines the area under the pulse, which will be constant, then it is seen that the average voltage output for any input voltage magnitude will be constant.

It is however, essential that the connection between the saturable type reactor and the rectifier be a nonreactive type of connection. This becomes apparent when the pulse output of the saturable type reactor to the reactive device is considered as comprising a series of harmonic voltages. Clearly, for different size pulses, which would be obtained with different input voltages, the harmonic content of the pulse would vary. If now a reactance carries this pulse of varying harmonic content,

it is obvious that each harmonic would see a different impedance. Therefore, a pulse of variable harmonic content would be attenuated according to the impedance seen by each constituent harmonic, and accordingly, the total output of the sum of each of the attenuated harmonies for different pulse shapes giving a different output voltage after rectification of, the pulse.

By meeting the restrictions of a connection which is substantially non-reactive, such as a direct or resistor connection, a selective impedance efiect upon the harmonics which comprise any particular pulse would not be affected and the output voltage pulse for any shape would indeed contain a constant area thereby afliording a constant D.-C. voltage output.

Accordingly, a primary object of my invention is to provide a novel circuit which will provide a constant D.-C. voltage output from a source of variable alternating voltage.

Another object of my invention is to provide a constant D.-C. voltage output circuit which utilizes the properties of a saturable type reactor.

A still further object of my invention is to provide a constant D.-C.' voltage output circuit which utilizes a saturable type reactor wherein the output voltage is determined as a function of the number of volt seconds which is absorbed by the saturable type reactor.

Another object of my invention is to provide a D.-C. output voltage circuit utilizing a saturable type reactor and a rectifier wherein the connection between the saturable type reactor and the rectifier is either a direct connection or a resistive connection.

In the practice of my novel invention as described above, it has been found that there is a very slight variation, which in many applications is negligible, of the D.-C. output voltage for variations in the A.-C. input voltage due to the air reactance of the saturable reactor winding when this reactor is saturated.

That is to say, that although the saturable reactor is in a saturated state a very small voltage will continuously be transmitted from the A.-C. input circuit to the rectifier since the saturated saturable core reactor acts like an air core transformer.

If the application is so critical that even this very small variation cannot be tolerated, I have found that it can be compensated for by providing a small auxiliary transformer in the A.-C. circuit which will induce a voltage into the circuit leading to the rectifier which substantially equally opposed the aforementioned voltage due to the air reactance of the saturable reactor winding for all values of input A.-C. voltage.

Accordingly, another object of my invention is to provide means to compensate for the voltages due to the air reactance of the saturable type reactor which are induced in the rectifier circuit.

Another object of my invention is to provide transformer means in the input A.-C. circuit which will induce a voltage in the circuit leading to the rectifier for bucking down voltage due to the air reactance of the saturable reactor.

Although the D.-C. output of my novel circuit is constant for any output pulse, it should be understood that the R.M.S. output and harmonic content varies Widely.

It is, however, necessary in some applications for the R.M.S. input to a load to be substantially constant.

I have found that I can accomplish this end by providing a filter for the output of the rectifier which comprises a high pass filter and a low pass filter. The load circuit, which must carry a direct current of constant magnitude, is connected through the low pass filter and a dummy load having an impedance equivalent to the impedance of the load is connected to the high pass filter.

The rectifier output impedance to the load will therefore be approximately resistive for a full range of harmonic frequencies and the operation of the circuit will not be affected by a frequency selective impedance of any intermediate component of the circuit and the load, in view of its low pass filter will carry a current of a substantially constant R.M.S. value.

Accordingly, another object of my invention is to pro- Patented Apr. 4, 1961 estates 3 vide a high pass filter andlow pass filter to the output of my novel circuit.

These and other objects of my invention will become pparent when taken-in conjunction with the figures in which; a i V F gure 1 is a circuit diagram of one form of my novel invention.

Figure 2 shows another form which can be taken by my novel invention.

Figure 3 shows a first, second and third voltage of the variable A.-C. voltage input as plotted against time.

Figure 4 shows the voltage appearing on the current l mmng impedance which is in series with the saturable reactor primary Winding.

Figure 5 shows the A.-C. voltage as a function of time which appears on the saturable type reactor winding as a voltage pulse'for the corresponding three voltages shown in Figure 3. a

Figure 6 shows the constant DC. voltage output of the rectifier for the three voltages of Figure 3.

Figure 7 shows the circuit of Figure 1 modified to contain means to compensate for the air core inductance of the saturable reactor and high and low band pass circuits for the rectifier output.

Figure 8 is similar to Figure 7 where the filter circuits are'slightly modified.

Referring now to Figure 1, an A.-C. voltage of variable magnitude is impressed upon the A.-C. terminals 10 and 11 and this voltage is then impressed upon the series connection of Winding 12 of the saturable reactor core 13 and (non-saturable) reactor 20. A second winding 14 is also wound in the core 13 of the saturable re- .actor to thereby allow a transformer action between the winding 12 andld when the core 13 is saturated. T he voltage appearing upon the coil 14 is then impressed directly upon a rectifier 15 and the output of this rectifier is then smoothed in any desired manner such as the filter comprising the resistor 21, the capacitor 16 and in- .ductor 17. As will be shown hereinafter, a DC. voltage of constant magnitude will appear on the output terminals 18 and 19.

- The circuit of Figure 1 is essentially duplicated in the circuit of Figure 2 with several illustrative variations. In Figure 2, the core 13 contains only a, single winding 12 which acts as both a secondary winding and a primary winding when the core '13 is unsaturated. Here again, a current limiting reactor 20 is provided. This is to say that whenever the core 13 is saturated, the current due to the alternating voltage impressed upon the terminals 10 and 11. will rise unhamper'ed if a current limitingdevice is not in the circuit. Obviously, the inductor 2% could have been replaced by another impedance such as a resistor but the choke action is more desirable since it dissipates less power.

Another modification shown in Figure 2 is in the positioning of resistor 21. Here, as in the case of Figure l, resistor 21 serves a current limiting function. Note,

The operation of the circuits of Figures 1 and'Z may now be described. in conjunction with the voltage time diagrams shown inFigures 3, 4, 5 and 6. More specificallyrFigure .3 .shows'a first voltage V, which is as sumed by the alternating voltage at the terminals 10 and 11,-a second and higher voltage V which-maybe the .voltage assumed at another time at the terminals 10 and 11 and a stillhigher voltagcv Wl ien the voltage V '0f Figure :3 appears across the l winding 12 of Figure l or 2, it is clear that the core 131 will remain unsaturated for the interval of time defined by the length of the pulse V shown in Figure 5.

Since the circuit is completely reactive, the current will be 90 out of phase with the voltage and therefore at both the beginning of the voltage cycle and at the end of the voltage cycle and the core 13 will be saturated, thereby requiring the shape of the pulse V to be substantially symmetrical about the peak of voltage V The area under the pulse V of Figure 5 is, of course, determined by the saturable reactor design which is so constructed as to absorb only a certain predetermined number of volt seconds. If now the A.-C. voltage is increased to achieve the value V of Figure 3, it is clear that in view of this higher voltage, and in view of the fact that the saturable reactor can absorb only a predetermined number of volt seconds,-that the voltage pulse V appearing on the winding 14 as shown in Figure 5 will be higher and, therefore, of a decreased duration. It is, however, essential to note that the voltage time area defined by the pulse is exactly equal to the area dofined by the pulse of voltage V of Figure 5.

It then follows that if the voltage is further increased to the value V of Figure 3, that the voltage V of Fig ure 5 which is of higher magnitude but shorter duration will appear on the saturable reactor winding.

During the interval that the saturable reactor is saturated, the input voltage, as shown in Figure 4, appears across the current limiting-reactor 29 of Figure l or- 2.

Hence, in the case of the voltage V of Figure 3, a very small voltage V appears on reactor 26 at the beginning of the cycle until the saturable reactor unsaturates. The A.-C. voltage then transfers to the much higher impedance of the unsaturated saturable reactor as shown by V of Figure 5. Upon the subsequent saturation of the saturable reactor, the voltage as shown in Figure 4 once again appears on the current limiting impedance 2% of Figures 1 and 2.

Similar remarks can be directed to the case of the input voltage V of Figure 3 on which the voltage V of Figure 4 appears on the current limiting impedance and the case of the voltage V in which the voltage V appears on this current limiting impedance.

It is important to note that in the case 0:. Figure l, the voltage pulse is transformed into the winding 14 so long as the core 13 is unsaturated and the voltage pulse is directlytransmitted to the input terminals or" rectifier 15 In the case of the circuit of Figure 2, the voltage pulse appearing on the winding 2. is transmitted to the rectifier input terminal through a resistor; this nonreactive type connection between the transducer winding and rectifier as was noted above, is essential, sinceit does not allow the selective impedance effect of a reactance to be exerted upon the transmitted pulse which is being put on the rectifier input terminal.

' After the rectifying action of the rectifier 15 of Figure 1 or 2, the voltage appearing at the DrC. terminals of the rectifier is as shown in Figure 6, the average D.-C. voltage being shown by the dot-dash line of constant magnitude. Note that the average voltage is constant whether it is due to the'short and wide pulse V the tall thin pulse V or the-intermediate pulse V This is due of course, to the-fact that the area contained under V age value is the integral ofthe voltage times the derivative of the time. This integral gives the area under the pulse. Therefore, two pulses having a constant area ,will,' of necessity, have a constantaverage value.

It is to be notedthat thejroot mean square value of voltage varies. Once again, this may be explained in the definition of the root means square value, and reference to the voltages shown in Figure 6. The root means square value of the voltages will be equal to the integral of the voltage squared times the derivitive of the time. Therefore, even though the pulses have a constant area, the diiference in the instantaneous voltage squared times the derivitive of the time summed up over the length of the pulse will obviously give completely difierent value for ditferent pulse shapes.

In the previous description of my novel invention, I have shown both Figures 1 and 2 as utilizing a bridge type rectifier. It is, however, clear that any rectifying means could be utilized in the operation of my novel invention. Similarly, the current limiting resistor 21 of Figure 2 is applied to the circuit of Figure 1, after the rectifier. This resistor is necessary to prevent an excessive inrush current through the rectifier 15 into the capacitor 16. Similar remarks can be directed to the use of a single winding 12 as shown in Figure 2 which would act as both a primary winding and secondary winding when the core 13 is unsaturated. That is, it is merely a design expedient which determines the use of a single winding or an isolated primary and secondary winding 12 and 14 as shown in Figure 1.

Figures 7 and 8 show two refinements of the original circuits in Figures 1 and 2. These circuits operate the same way as the other circuits, except that they have two compensations to provide a more nearly constant D.-C. voltage over a wide variation of the A.C. voltage.

In Figures 7 and 8, the choke 24 replaces the choke 20 of Figures 1 and 2. This choke absorbs the voltage shown in Figure 4, on the winding 22 which is in series with the saturable reactor winding 12. In the winding 23 on a low voltage proportional to the one on winding 22 is induced. In the windings l2 and 14 a voltage is induced according to Figure 5. However, the volt-time area of this voltage is not exactly constant because the mutual air-core inductance of the coils 12 and 14 induces a small additional voltage in coil 14; this additional voltage being proportional to the current in the coil 12. Since the voltage on coil 22 increases greatly when the voltage on the terminals -11 increases, which also causes the current to increase, the undesirable additional voltage in coil 14 is approximately proportional to the voltage on coil 22. As the circuits in Figures 7 and 8 are connected, the small voltage induced in coil 23 subtracts from the voltage induced in coil 14 which compensates for the additional voltage caused by the air core inductance.

Another variation over Figures 1 and 2 is in the filter. In Figure 7 the filter consists of a high-pass filter consisting of two capacitors 25 and 26 and a reactor 27. This filter is terminated in a resistor 28 which has the same resistance as the load 32. the rectifier is not distorted (V in Figure 6) the high frequency components of this voltage are small and the current absorbed by this high-pass circuit is very low. When the voltage output of the rectifier is greatly distorted (V in Figure 6) then the high-pass circuit carries a high harmonic current which is absorbed by the resistor 28.

The load circuit, which must carry a direct current of constant magnitude, is connected through a low pass filter consisting of the reactors 29 and 31 and the capacitor 30. The reactor 31 is not necessarily a real coil but is usually represented by the inherent inductance of the D.-C. load. This low-pass circuit keeps the current through the load 35 at a constant value. I

If the voltage output of r By adding a low-pass filter to the load and a highpass filter to a dummy resistor 28 we obtain a rectifier output impedance which is constant and approximately resistive for the full range of frequencies from D.-C. to infinite. Therefore, the operation of the constant voltage circuit is not affected by a frequency selective impedance of any interval component of the circuit.

In Figure 8 the circuit is made according to the same principle, except that the high-pass and low-pass filters are more elaborate: The high-pass filter consists of the capacitors 25, 26 and the inductance-capacitance circuit 2733. The low-pass filter consists of the reactors 21 and 31 and the inductance-capacitance circuit 3430. This has the advantage that the combined impedance as seen from the rectifier terminals is more nearly and constant for the full range of frequencies.

In the foregoing, I have described my invention only in connection with preferred embodiments thereof. Many variations and modifications of the principles of my invention within the scope of the description herein are obvious. Accordingly, I prefer to be bound not by the specific disclosure herein but only by the appending claims.

I claim:

1. A circuit to supply a constant D.-C. output voltage from a variable A.C. input; said circuit comprising a saturable type reactor and a rectifier; said saturable type reactor having a primary and a secondary winding and a core of rectangular hysteresis loop material; said secondary winding directly connected to the A.C. side of said rectifier and said primary winding connected in series with said variable A.C. source and filtering means being connected to the D.-C. terminals of said rectifier, said filter means comprising a high pass filter and a low pass filter, said low pass filter being connectable to a DC. load, said high pass filter being connectable to a dummy load.

2. A circuit to supply a constant D.-C. output voltage from a variable A.-C. input; said circuit comprising a saturable type reactor and a rectifier; said saturable type reactor having a primary and a secondary winding and a core of rectangular hysteresis loop material; said secondary winding directly connected to the A.C. side of said rectifier and said primary winding connected in series with said variable A.C. source and filtering means being connected to the D.-C. terminals of said rectifier, said filter means comprising a high pass filter and a low pass filter, said low pass filter being connectable to a D.-C. load, said high pass filter being connectable to a dummy load and transformer means being connected to induce a compensating voltage from the circuit including said primary winding to the circuit including said secondary Winding, said compensating voltage being opposed to the voltage induced in said secondary winding due to the air core reactance of said saturable type reactor.

References Cited in the file of this patent UNITED STATES PATENTS 1,960,599 Silva May 29, 1934 2,084,870 Schmidt June 22, 1937 2,341,446 Klinkhamer et al. Feb. 8, 1944 2,442,960 Pohm June 8, 1948 2,725,515 Horton Nov. 29, 1955 2,739,282 Evans Mar. 20, 1956 2,763,827 Evans Sept. 18, 1956 2,908,864 Shepard Oct. 13, 1959 FOREIGN PATENTS 470,351 Great Britain Aug. 13, 1937

Images