Patent Grant
9791878
Field of Invention
The present invention relates generally to electrical circuits for voltage control, and more particularly, to digital circuits for voltage control, and hence delivery of power, to electrical loads.
Description of the Related Art
Electrical devices and appliances are generally designed to operate at specific power supplies in terms of voltage magnitude and frequency, and other properties. When in use, any deviation from the specified powering conditions could render the devices or appliances inefficient, inoperative or even permanently defective.
Therefore, ever since human deployment of electricity, it is a common goal of electrical and electronics engineers and scientists to develop devices and methods to control and deliver electrical power to the loads efficiently. Various inverters, converters, voltage regulators, power amplifying and power switching components, electrical sensors, etc. are invented and developed to increase the capabilities in electrical power control, in terms of energy efficiency, control accuracy, response speed, power level and system cost, etc.
In practice, electrical voltage control techniques are employed to control electrical parameters other than voltages. For example, by controlling the voltage applied to a constant impedance device, current through the device is controlled. As another example, by controlling the voltage applied to a load, the power generated by the load is controlled. Thus in practice electrical voltage control apparatus functions in many different forms and in many different application areas, such as:
Broadly speaking, we may define two distinct methods in voltage control, namely analogue and digital approaches. By analogue approach, voltages are scaled up or down continuously through any voltage levels within the control range. By digital approach, to be called Digital Voltage Control throughout this specification, voltage levels are “stepped” through discrete levels within the control range. One common way is, as often used for AC mains regulation, through switching in and out of transformer coils as “voltage cells”, i.e. voltage sources which are in galvanic-isolation from each other.
There are a number of merits of digital approach over the analogue approach. The analogue approach is through linear control of active electronic devices such as transistors operating in the linear mode, or through circuit switching by active electronic devices such as transistors operating in the switching mode. By linear mode of operation, voltage control can be achieved with high control accuracy and high control speed but at the cost of low power efficiency. By switching mode of operation, voltage control can be achieved with high power efficiency but often with compromised control accuracy and control speed. For very high power applications, the analogue approach, either in linear or in switching mode, faces the difficulties of very high cost or unavailability of suitable active high power or high frequency devices. Further there are more EMI and EMC issues in association with high power and high frequency switching.
The digital approach as adopted by present invention is through switching in and out of “voltage cells” at the usually low power frequency (such as the 50 or 60 Hz mains frequency, or even DC) of the voltage under control, rather than at very high frequencies. Demand on switching speed of the switching devices as well as on the control schemes are not high in general, even at very high power levels. Further, as switching is performed at low frequencies, the issue on EMI or EMC is relatively less serious and might be more easily handled. The digital approach is therefore a better choice to the analogue approach when power-handling capacity and low cost are the prime considerations. Moreover, since the switching loss at low frequency is relatively low, the digital approach enjoys also the benefits of high power efficiency. Furthermore, that no and little distortion is introduced through switching is yet another advantage by the digital approach as compared to the analogue approach.
However, there is still a very important aspect of voltage control or regulation to be considered, namely the accuracy of control. Since by the digital approach, the voltage is varied by steps, the accuracy of control is always limited by the size of the voltage steps. It is obvious for a fixed range of voltage control, the fineness of control is inversely proportional to the number of voltage levels that could be “stepped” through. It is also obvious that for a fixed number of voltage levels, all voltage steps should be made equal to achieve the highest accuracy of control.
When the number of steps is increased for the purpose of achieving finer control, the number of switches required will inevitably increase. Since the switches are the key and relatively expensive components of the system, accuracy of control has often been compromised for lowering the system cost by limiting the number of switches deployed. This is highly undesirable and many different varieties of switching circuit topologies and control methods have been attempted in the past to achieve higher control accuracy while limiting the number of switches employed for circuit simplicity and cost reduction. However these existing designs are in general complicated in overall system structure, restrictive in deployment and often overly complicated in control methodology.
Further, when fine steps are achieved for high control accuracy, a new challenge of maintaining system stability will be in front of the designer. Dependent of the actual circuit design and the accuracy in circuit implementation, monotonicity between the digital control signal and the controlled step voltage output would be lost as the size of the steps decreases to some extent. Consequently, lack of monotonicity causes system instability and also reduction in control accuracy.
While piecemeal improvements or alterations are revealed in many prior inventions, none has actually proposed a unified approach to address the above issues. The present invention is intended to solve all these problems and it will become clear when the invention is disclosed herewith exemplary embodiments.
Prior arts in the voltage control or regulation are found typically in AC voltage regulators, whereby many methods and devices are developed to control the AC voltage through digital approach, and some are revealed by the following patents:
In majority of the above inventions and disclosed embodiments, the circuit topologies proposed tend to be very specific and hence very restrictive. The restrictiveness in circuit topologies has presented difficulties to the designer in optimizing the performance of the voltage regulator under practical considerations, such as the difficulty in deciding the best number of voltage modules, the best number of voltage cells in each voltage module (such as the number and turns of transformer coils in the design of transformers for tap-switching voltage regulators), the best number of switches in each voltage module, the most suitable control methodologies and control modules, etc. Consequently, there is a lack of design flexibility for optimizing the performance of the voltage regulator in terms of accuracy of control, voltage range of control, speed of response, cost of implementation, and cost of maintenance, etc.
Further, linearity and monotonicity of the voltage variation are not generally addressed. In many of the inventions, the equal voltage steps are not achieved or not even intended to be achieved. The step sizes are simply not constant by design in these inventions. The result is that the voltage change is non-linear or even worse, not monotonic. Non-linearity will lower the control accuracy achievable, while non-monotonicity will render a feedback control system unstable. Both are detrimental to the performance of the digital voltage control system.
Further still, none of the prior inventions has addressed the issues on the practical limitations affecting the linearity and monotonicity of the voltage under digital control. Consequently the performance of the digital voltage control system, in terms of control accuracy and system stability, is likely compromised due to the oversight of this aspect in system design.
In most cases, prior art designs fail to show the ideal or the preferred theoretical ratios of the voltage cells. In a number of cases, some ratios are proposed without any reasoning as how these ratios are arrived at. Consequently there is no guidance in design to optimize the system, in terms of control accuracy and control range, through proper selection by design the number and magnitude of the voltage cells, and the voltage ratios between the voltage cells.
As will be clear from the following detailed description, the present invention adopts a unified approach to address the above issues not sufficiently addressed before. Apart from practical limitations of components available, there is no restriction by the present disclosed approach in designing the voltage controller in terms of control accuracy, number of switches deployed, number and magnitude of voltage cells. The method of control and the associated control circuitry is simple and straight forward, while the practical limitations affecting the linearity and monotonicity will be addressed to have its consequent bad effects removed too. This will be explained in details with disclosed embodiments for illustration.
Despite that above quoted prior arts dual with transformer tap-switching voltage regulators or controllers, whereby independent transformer coils are depicted as voltage cells in galvanic-isolation, the present invention applies to any other electrical voltage sources in any forms (named “voltage cells” throughout this patent specification) such as these quoted below as examples:
Furthermore, by the duality property of electrical circuits, the present invention can be applied also to electrical current control, as will be explained in more details.
The objective of the present invention is to provide the apparatus and method of controlling the output of a voltage supply, through digital circuit switching of a combination of voltage sources, to vary in a series of voltage steps accurately and speedily within a predetermined control range. It is a design target to have all the steps of equal magnitudes. Any deviation in practice from the target would need to be kept within a maximum value such that any of the controlled change of the output of the voltage supply within the control range is always monotonic, i.e. the output of the voltage supply always increases or remains constant as the value representing the digital input increases, while the output of the voltage supply always decreases or remains constant as the value representing the digital input decreases. Monotonicity is an essential characteristic for stability within a control system.
With the foregoing in view, as other advantages as will become apparent to those skilled in the art to which this invention relates as this patent specification proceeds, the invention is herein described by reference to the accompanying drawings forming a part hereof, which includes descriptions of typical embodiments of the principles of the present invention, in which:
Glossary
The present invention falls into the specific area of Digital Voltage Control. In
As shown in
Within each of the Voltage Modules VMs, there are one or more Voltage Cells VCs as voltage sources connected in series aiding, the Voltage Cells VCs being in galvanic isolation before said connection, the total number of VCs in the mth Voltage Module VMs being Nm, where m=1 to MVM.
Within each of the Voltage Modules VMs, all the Voltage Cells VCs provide voltages identical in magnitude, in waveform and in phase. In other words, The nominal voltages of the Voltage Cells VCs Vm1=Vm2=Vm3 . . . =Vmc where c=1 to Nm, for any of the Voltage Module VM. The value of the identical voltage is designated Vm as the nominal voltage in the mth Voltage Module VMm.
Within each of the Voltage Modules VMs, all the Voltage Cells VCs are connected in series such that each cell is adding to the overall voltage of the Voltage Module VM. In other words, the highest voltage achievable from the Voltage Module VM is the direct sum of voltages from all the Voltage Cells VCs within the Voltage Module VM, i.e. Vm.Nm in the mth Voltage Module VM.
Within each Voltage Module VM, two sets of switches in Series-S and Series-J respectively, and designated by S1, S2, S3, . . . Si . . . where i=1 to Nm+1, and J1, J2, J3, . . . Ji . . . where i=1 to Nm+1 respectively, and connected in parallel with the series of Voltage Cells VCs such that at any time each of the two output terminals To1 and To2 of the Voltage Module VM can be connected to any one of the connection nodes T1, T2, T3 . . . Ti where i=1 to Nm+1 within the Voltage Module VM.
For all Voltage Cells VCs in all Voltage Modules VMs, i.e. VCmc, for cth VC in mth VM, for c=1 to Nm, m=1 to MVM, the nominal voltages Vmc are identical in waveform and in phase but are of different magnitudes between the Voltage Modules VMs. However the ratios of the magnitudes of voltage Vm of the voltage cells between the Voltage Modules are uniquely defined by present invention, and according to the number of Voltage Modules and numbers of Voltage Cells VCs in each of all the Voltage Modules as V1:Vm=1:Π(Ni+1) where i=1 to m−1, m=1 to MVM (Π denoting the product of the series).
Further by turning on the appropriate switches of the S-series and J-series, the output of the mth Voltage Module VMm, i.e. Vmo, can be varied monotonically from −VmNm to +VmNm in practically equal steps of Vm.
By combining all the Voltage Modules, and by turning on the appropriate switches, the output of the Digital Voltage Controller VDVC can be varied monotonically from −ΣVmNm to +ΣVmNm in practically equal steps of V1.
It is possible by suitable design all voltage cells are in series aiding so that the resultant current always flows in one direction. However by the combined switching actions of the S-series and J-series switches the output from a Voltage Module can vary between opposite polarities. Consequently in combination with other Voltage Modules which might also have either polarities, it can be expected that some of the voltage cells need to handle currents flowing in either directions. In that case, these voltage cells would need to be made to allow bidirectional current flow, and hence bidirectional power flow.
There is another practical issue to be considered. Due to various reasons, such as the tolerance in physical implementation of the Voltage Cells, partly due to the uncertainty in voltage measurements, the loading effect on the Voltage Cells whenever load current is drawn from the Digital Voltage Controller DVC at VDVC, and the voltage drop on the switches, etc., there is likely some departure in the contribution of Vmc from the nominal value.
Designating this departure by δVmc, and assuming δVa is the largest voltage deviation for any of ALL Voltage Cells VCs in all Voltage Modules VMs, i.e. the largest among all δVmc, c=1 to Nm, m=1 to MVM, the maximum departure of Vmo, voltage output from mth Voltage Module VMm is δVmo, then δVmo=δVa×Nm.
Therefore, in comparison to the nominal voltage values, the deviation of VDVC, the output of the Digital Voltage Controller DVC, is δVDVC, and the maximum value of δVDVC=ΣδVmo=ΣδVa×Nm=δVa×ΣNm where m=1 to MVM.
Note that δVmc, δVa, δVmo, and δVDVC can be of either positive or negative values. However for the estimate of their largest possible values, we shall take from here their absolute values, i.e. the magnitudes only.
If the steps are controlled such that each time one and only one Voltage Cell VC in any of the Voltage Modules VMs is added to or removed from contributing to the output of the Voltage Module VM, the maximum change in the voltage deviation from the nominal, at each voltage step, for the Voltage Module would be δVa, and total change in the voltage deviation from the nominal voltage VDVC for the Digital Voltage Controller, is δVDVC=|δVa|×M
In order to make sure that the voltage change under control is monotonic, it is required that |δVDVC|<V1 i.e. |δVa|×MVM<V1, i.e. |δVa|<V1/MVM
In other words, the deviation of voltage of any of the Voltage Cells VCs would need to be less than the nominal voltage of the smallest (least significant) Voltage Cell divided by the total number of Voltage Modules VMs.
Hence to maintain monotonicity in practice the voltages of the Voltage Cells VCs between the Voltage Modules VMs are allowed to bear the ratios V1:Vm=1:[Π(Ni+1)]±|δVa|/V1=1:[Π(Ni+1)]±1/MVM where i=1 to m−1, for m=1 to MVM.
Shown in
On the other hand, Series-J switches in the Voltage Modules VM1, VM2 and VM3 are manually set through control signal lines of CSJ1, CSJ2 and CSJ3 respectively to shift the output voltage from each of the Voltage Modules VM1, VM2 and VM3 as required by the design. In the case the Series-J switches are to be permanently set, these switches can be replaced by hard-wiring, i.e. shorting for closed switches and opening for open switches, within the Voltage Modules VMs as will be shown by an exemplary embodiment of the invention to be described next with reference to
In operation, when Vreg is lower than the set-point by a pre-defined amount the Counter Control CC will trigger the cascaded counters to count-up so that VDVC or Vreg is raised. When Vreg is higher than the set-point by a pre-defined amount the Counter Control CC will trigger the cascaded counters to count-down so that VDVC or Vreg is lowered. Vreg is thus controlled to a value close to the set-point despite of any variations in the supply voltage at the input of the voltage regulator or any variations of the load at the output of the voltage regulator.
Referring to
The Digital Voltage Controller DVC basically consists of a multi-coil multi-tapped power transformer T1. As shown there are three transformer coils acting as Voltage Modules VMs designated by VM1, VM2 and VM3, respectively. Here MVM, the total number of Voltage Modules VMs is therefore 3. VM1 has 2 segments of the corresponding transformer coil and therefore 2 Voltage Cells VCs, similarly VM2 has 3 Voltage Cells VCs while VM3 has 4 Voltage Cells VCs. Each Voltage Cell corresponds to a segment of the transformer coil, each segment bearing the same number of windings within each coil. The Voltage Cells bear the voltage ratios 1:3±0.33:12±0.33 between the Voltage Modules VM1, VM2 and VM3. The transformer taps are switched on and off by Series-S switches under the digital control signals from the Control Module CM.
The 3 taps of coil VM1 are connected to switches by S11, S12 and S13 respectively. The 4 taps of coil VM2 are connected to switches S21, S22 and S23 and S24 respectively. The 5 taps of coil VM3 are connected to switches S31, S32 and S33, S34 and S35 respectively.
As shown in
As shown by the table, by controlling the switch states of the 3 sets of Series-S switches in the 3 Voltage Modules VMs for a total of 60 states, the voltage output from the DVC can be stepped from −30 to +29 in steps of one, each step being 1/48 of the regulator output voltage, i.e. Vreg/48.
Referring back to
The Series-S switches are driven by the counters which are designated as
CS1 for VM1: Ternary (Modulo-3) up-down counter
CS2 for VM2: Quaternary (Modulo-4) up-down counter
CS3 for VM3: Quinary (Modulo-5) up-down counter
The three counters are cascaded to count through a total of 5×4×3=60 counting states.
The polarity of compensation transformer T2 is chosen such that when Vreg is lower than the set-point by a pre-defined amount the Counter Control CC will trigger the cascaded counters to count-up so that an increasing VDVC is generated and added through the secondary side of the compensation transformer T2 resulting a higher Vreg. Similarly when Vreg is higher than the set-point by a pre-defined amount the Counter Control CC will trigger the cascaded counters to count-down so that a decreasing VDVC is generated and added through the secondary side of the compensation transformer T2 resulting a lower Vreg. Vreg is thus controlled to a value close to the set-point despite of any variations in the supply voltage at the power input of the AC voltage regulator or any variations of the load at the output of the AC voltage regulator.
The scale of compensation is dependent on K, turns ratio of the primary to the secondary of the compensation transformer T2. A larger K means a smaller scale of compensation, and a smaller K means a larger scale of compensation. As an example, for a nominal Vreg=220 volt, K=6, theoretically the AC voltage regulator will be able to maintain an output accuracy of ±0.34% for supply voltage variation in the range ±20%. Similarly, for K=3, the AC voltage regulator will be able to maintain an output accuracy of ±0.68% for supply voltage variation in the range ±40%.
Note that Series-J switches are not deployed in the above embodiment. As shown in
Note also that there is much more room for variation in the embodiments of the present invention. Both the number of Voltage Modules VMs and the number of Voltage Cells VCs in each Voltage Module VM can be chosen to suit individual design considerations, such as control accuracy, control range, number of transformer coils (the Voltage Cells VCs and Voltage Modules VMs) and number of switches needed, control circuit complexity, system stability, total implementation cost, etc. Room of variation available in the present invention provides much design flexibility in choosing the best circuit topology.
By duality property of electrical circuits, all circuit principles described for voltage control in this application can be applied also to current control. As an exemplary embodiment of present invention the basic architecture of a Digital Current Converter DIC is shown in
Although the present invention and its advantages have been described in detail, it should be understood that various changes, substitutions, and alterations can be made therein without departing from the spirit and scope of the invention as described. That is, the discussion included in this application is intended to serve as a basic description. It should be understood that the specific discussion may not explicitly describe all embodiments possible; many alternatives are implicit. It also may not fully explain the generic nature of the invention and may not explicitly show how each feature or element can actually be representative of a broader function or of a great variety of alternative or equivalent elements. Again, these are implicitly included in this disclosure. Where the invention is described in device-oriented terminology, each element of the device implicitly performs a function. Neither the description nor the terminology is intended to limit the scope of the invention.
| Filing Document | Filing Date | Country | Kind |
|---|---|---|---|
| PCT/IB2013/056337 | 8/2/2013 | WO | 00 |
| Publishing Document | Publishing Date | Country | Kind |
|---|---|---|---|
| WO2014/020572 | 2/6/2014 | WO | A |
| Number | Name | Date | Kind |
|---|---|---|---|
| 5710504 | Pascual | Jan 1998 | A |
| 20120161526 | Huang | Jun 2012 | A1 |
| 20140312700 | Catthoor | Oct 2014 | A1 |
| Number | Date | Country | |
|---|---|---|---|
| 20150241900 A1 | Aug 2015 | US |
| Number | Date | Country | |
|---|---|---|---|
| 61678683 | Aug 2012 | US |
