Current measuring device for a power converter and regulating circuit for applying the same

Abstract

The invention relates to a current measuring device for a clock-controlled power converter, having two independent sensors connected in series for the current to be measured, where one sensor is a susceptance and the other is a conductance. A parallel current measuring resistor and an averaging capacitor provide a share for the susceptance. The conductance is formed by the series-connected current measuring resistor. The current to be measured has an effective value which is at least 1.4 times and at most 8 times as large as the average value. The current has a DC and an AC current component. For regulating the power converter, the evaluation of the DC current component or the average value is particularly important. Different from the prior art in which the input of an electrical filter is designed to be as highly resistive as possible for measurement purposes, the filter is shifted back into the power path or fused with the current measuring sensor. For this purpose, the known centralized current sensors from the prior art have to be divided. The pre-filtered measurement voltage can be evaluated more precisely, especially when the evaluation is divided similarly to the measurement.

Description

Field of the Invention

[0001] The present invention relates to a current measuring device for a current having a high alternating current component, which current occurs, for example, particularly in clock-controlled electronic power converters. Due to this alternating current component, the measurement must be filtered. The invention also relates to a regulating circuit for such a clock-controlled power converter. Clock-controlled electronic power converters are particularly included in plug-in power supplies, simple power supply devices, switched-mode power supplies, laboratory power supplies, DC current converters, DC voltage converters, active current sources, active voltage sources or operating devices for light sources, preferably light-emitting diodes. Background Art

[0002] Today, most electronic circuits for converting electrical energy or power operate in a clock-controlled manner and almost always have to be regulated. For this purpose, within such a clock-controlled electronic power converter, an electrical signal that is most effective for at least one parameter to be regulated must be measured in order to be able to report the measurement value thus obtained as a feedback parameter to the regulator. For example, if the power is to be regulated to its rated value, then at least one signal representing the voltage causing the considered power must be measured, and additionally at least one second signal representing the current causing the same power must be measured.

[0003] However, usually only one or the other signal is sufficient. If the power converter operates as an active voltage source or a DC voltage converter, then only the signal representing its output voltage is processed in its regulation. Conversely, if only the signal representing the output current is processed there, then the power converter operates as an active current source or a DC current converter. The respectively open, i.e., non-regulated parameter is defined by the electrical load that is exactly connected to the considered power converter.

[0004] However, in most cases, signals representing such open parameters are measured together. In this way, the voltage source or DC voltage converter is protected against overload, which intervenes in the case of the maximum allowable load current and converts the actual voltage source into a current source, thereby preventing the load current from being greater than the maximum allowable load current. A simple power supply device constructed in this way, such as a plug-in power supply, is regulated to its output voltage during normal operation, contrary to its name. In this way, the active current source or DC current converter obtains no-load stability, which is particularly important for a power converter that should generate a so-called safety extra low voltage (SELV). Since there is no load, the current source would produce an infinitely high output voltage. When measuring the output voltage together, this output voltage can be very effectively limited to a maximum allowable value, such as 54 V or 60 V. This operating mode of "current source characteristic with voltage limitation" is one of the most common in operating devices for light-emitting diodes.

[0005] A more general level is the so-called laboratory power supply, which can perform the above two tasks according to the load and current settings and is mainly constructed as a switched-mode power supply, i.e., as an energy supply device that includes at least one clock-controlled electronic power converter. Usually, two regulating circuits are arranged in parallel with each other there, where one regulating circuit with more stringent rating limits is always in intervention and regulates the power converter. The physically open parameters also remain open during regulation. Or these two regulations are cascaded with each other, where the current regulation is mostly subordinate. This is because the current absorber (Stromsenke) dynamically invalidates the always-present output capacitor due to its voltage application, so the current regulation is more easily stabilized than the voltage regulation. Both arrangements (at least in theory) result in a rectangular operating range in the IU diagram, which is spanned by the load current as the x-axis and the output voltage as the y-axis. The maximum power must be able to be handled by the power converter and can only be retrieved at one point, namely at the "upper right corner" of the operating range with the maximum load current and simultaneously the maximum output voltage, which leads to a very poor utilization of the components of the clock-controlled electronic power converter provided for this purpose.

[0006] All clock-controlled electronic power converters each include at least one storage capacitor and at least one storage inductor or storage coil or storage choke and at least one rectifying diode and at least one actively controllable power transistor. Thus, three of the simplest power converter topologies have been outlined: the buck converter, the boost converter, and the buck-boost converter (Buck-Boost or Flyback). The buck-boost converter has two storage capacitors. Alternatively, if two storage inductors that can also be coupled are provided and the storage capacitors are also moved between the power transistor and the rectifying diode, then a converter can be constructed therefrom. The Zeta converter and the SEPIC (single-ended primary inductor converter), i.e., the Zeta converter at its input and the SEPIC at its output, starting from a converter also require a second storage capacitor. The current converter (forward), which seems to be derived only from the buck converter, in addition to this simplest converter also requires two additional diodes and a three-winding transformer. If two actively controllable power transistors are used in the same power converter topology, then an improved current converter, an improved buck-boost converter, or a combination thereof, a synchronous variant of the six simple converter topologies from the buck converter to the SEPIC described above, or its associated bidirectional converter, push-pull circuit, or half-bridge is formed. The half-bridge, in addition to at least one additional rectifying diode, also requires at least one additional storage capacitor connected in series with the storage inductor, and this additional storage capacitor absorbs the DC voltage component of the half-bridge output voltage. If this series capacitor, together with the storage inductor, which in the case of the half-bridge is advantageously at least partially implemented by a transformer, forms an oscillating circuit having a natural frequency just below the lowest clock frequency, then a resonant LLC converter is produced. If the said transformer is additionally connected in parallel with a pure resonant capacitor, then a so-called resonant LCC converter is produced.

[0007] All of these are also usually power converters in the literal sense: The power converter can operate at the "upper right corner" above, but without significant changes to its circuit, the power converter can very well provide twice the voltage and only half the load current for this purpose, or twice the load current at only half the output voltage. The resulting hyperbola of constant power significantly increases the possible operating range and thus significantly increases the utilization of the clock-controlled electronic power converter within the considered power supply device. However, for this purpose, as the third point already mentioned at the beginning, power regulation is required, which acts as a hybrid between voltage regulation and current regulation. The measured values are multiplied, and if the product exceeds a preset power or the maximum power, then the third parallel regulation circuit intervenes, which regulates the power converter to exactly the preset power or its maximum allowed output power. As an alternative to this, the voltage rating or current rating can be correspondingly reduced, especially when the hyperbola of constant power can be approximated by a falling straight line crossing it, or when there is cascade regulation. A large number of modern operating devices for light-emitting diodes in all lighting fields are constructed in this way and can therefore also operate along the hyperbola of the same or maximum power. Thus, a large number of different loads, such as light-emitting diode modules, can be alternately powered by the same operating device without significantly increasing costs. In addition, all systems formed in this way can be dimmed almost over the entire brightness range, i.e., from 100% brightness to <1% brightness, for example, via communication protocols such as DALI, DMX, KNX, EIB, or RS488. Of course, wireless communication protocols such as Bluetooth, ZigBee, or Thread can also be used. For this reason, the operating devices suitable for such systems and the clock-controlled electronic power converters installed therein must cover a particularly large operating range, which approaches or even touches the y-axis of the IU diagram above. If the clock-controlled electronic power converter used in the operating device (especially when it is constructed resonantly) has natural operating boundaries, then it can also be used to limit power by simultaneously omitting the current regulation circuit and the voltage regulation circuit. Nevertheless, in all power converters for operating devices, laboratory power supplies, switching power supplies, DC voltage or DC current converters, active voltage sources or current sources, etc., at least two substantially independent electrical parameters must be continuously measured.

[0008] However, this "continuous" contradicts the "clock-controlled" that characterizes the operation of the considered power converter mentioned at the beginning. In order to achieve the desired effect of this clock control, first, one of the power converter topologies mentioned above is required. Secondly, due to the clock control, the control of at least one actively controllable power transistor included in the power converter topology is in principle pulse-width modulated (PWM), that is, it shows a period duration T, an on-time, and an off-time. The sum of the on-time and the off-time yields the period duration, or generally speaking, one of the three times is always related to the other two. This time segmentation, that is, all signal forms caused by the clock control, such as zero pauses, edges, slopes, peaks, etc., and all relevant time points or time intervals are the characteristics that most clearly define the clock-controlled operation of the power converter. Thirdly, there, each power transistor either always operates only fully conducting and controlled in saturation, or is fully cut off. Here, the term "power converter" always also includes DC current converters, DC voltage converters, switching power supplies, power supply devices, operating devices for LEDs, etc.

[0009] The further minimization of losses in the power converter is the motivation for this proposed current measuring device for a clock-controlled power converter. Because losses occur not only in the power converter topology itself, but also when measuring its instantaneous electrical parameters, especially when measuring current. For reasons of cost, speed, accuracy, and anti-aging, the current to be measured is often led through a current measuring resistor, and the non-infinite conductance of this current measuring resistor generates a voltage that can be measured across this current measuring resistor and used as a signal for the current to be measured. The value of this current measuring resistor is calculated by dividing the measurement voltage required for evaluation by the maximum current to be measured, and the cheaper or less precise the evaluation circuit set, or the more precise the evaluation must be, the higher this value is. Because a higher measurement voltage is required. The latter is the incentive for the present invention, which includes a second calculation rule for the current measuring resistor, which is particularly related to the left edge of the operating range of the power converter: the voltage resolution of the evaluation circuit (e.g., μV / bit) divided by the current difference required for the smallest measurable or smallest visible brightness difference for, for example, an LED operating device, which may result in a significantly higher value and higher losses for the current measuring resistor. The resulting second contradiction is "measurement resolution or accuracy" and "measurement loss", where the resolution at the left edge of the operating range (e.g., at a 1% dimming level) is preset by the DALI standard among other things, and must be particularly fine there based on the logarithmic eye characteristic curve. For conventional evaluation circuits and known solutions, this increases the measurement loss or the measurement and evaluation overhead or forces an increase in both at the same time.

[0010] The first contradiction between "continuous" and "clock-controlled" leads to additional particularities in the power converters considered here. Its regulation requires the continuity of measurement, but the measured signal is generally never continuous based on clock control. Therefore, in principle, the signal must be filtered before it can be fed to the regulation. The filter required for this purpose can also be included in the regulator or its compensation circuit or negative feedback circuit. Here, the boundary frequency of the regulated power converter, i.e., the frequency of the assumed modulation in the rated value of the output parameter, for example, is always significantly lower than its clock frequency, and the output parameter can be transmitted by the regulated power converter to its output without attenuation substantially. The clock frequency corresponds to the reciprocal of the above-mentioned period duration T of the pulse width modulation, and is usually at least ten times this boundary frequency. The time constant of the filter for the measurement signal also affects this boundary frequency, and this filter is mainly a low-pass filter. The entire system composed of measurement, regulation, clock control, and thus the actual energy transfer in the power converter topology is always much slower dynamically than the clock control itself.

[0011] Because in each power converter topology, in addition to the power transistor that chops the continuous region and the rectifier diode that generates another continuous region from this chop, there are basically only reactive elements (which operate with almost no losses and form an average value in principle due to their integrated characteristics), the calculation of these continuous regions allows for the use of the average value method, which is also known in the literature as "state space averaging": the average value of each chop signal also represents the signal of the continuous region, usually even only distinguished by a constant factor (usually 2 or 4), or by a factor related to the duty cycle of the above-mentioned pulse width modulation (PWM), or by other adjacent factors. The duty cycle is the ratio between the conduction time and the period duration within the exactly considered clock control period. Each low-pass filter is the technical counterpart of average value formation, and thus usually even automatically provides a result representing the continuous region while taking the duty cycle into account.

[0012] The advantage of directly measuring the chopping parameters (i.e., measuring the time curve or instantaneous value of a parameter, i.e., current or voltage, without significant low-pass filtering) is their additional information content regarding time segments. In particular, specific events, such as reaching the maximum current of a power transistor, can be associated with discrete time points, which can be directly translated into corresponding reactive control actions (such as overcurrent protection of the transistor or switching off the transistor within each clock control cycle). Thus, the measurement of such chopping parameters facilitates a return path not described hitherto, to re-form, in a clock-controlled manner, the control for at least one power transistor based on the continuous regulator output signal. The same measured signal after low-pass filtering can be fed, for example as a feedback parameter, to the input of the relevant regulator at the same time.

[0013] The measurement of the chopping parameters is in itself particularly important in those power converters which, in addition to the variation of the voltage and current values between the input and output, must additionally provide current isolation between their input and output. All power converters with the already mentioned safety extra-low voltage (SELV) make this an essential requirement, so that their output lines are allowed to remain touchable even during operation. In order to still be able to transfer power across the current barrier, a transformer can be integrated into the power converter topology, which is located, in the direction of the power flow, between the (multiple) power transistors and the (multiple) rectifier diodes. Thereby, each such topology is broken down into a primary side and a secondary side, the primary side having an input, at least one power transistor and at least one primary winding on the transformer, and the secondary side having at least one secondary winding on the same transformer, at least one rectifier diode and an output. The smallest topology just for this purpose (and thus also the reason for its popularity) is the flyback converter, whose input and output are connected in parallel to storage capacitors respectively. The primary and secondary currents in the transformer can have a DC current component, whereas, conversely, the primary and secondary voltages are pure AC parameters. Thus, all four parameters contain the main AC components to be measured all together.

[0014] In these insulated or isolated power converters, it is difficult that the power transistors have to be controlled on the primary side, while the important parameters, namely the load current and the output voltage, are present on the secondary side. Since it is particularly advantageous for the regulation and control of the power transistors to take place on the same side (i.e., the primary side), the measurements should also be carried out there. The same transformer that transfers the power to the secondary side can simultaneously transfer some of the parameters to be measured back from there to the primary side, which parameters, however, are mostly pure AC parameters, i.e., chopping parameters. During the off-time of the power transistor, the transformer primary voltage usually represents the current output voltage, while the current flowing through the conducting power transistor, i.e., usually the transformer primary current, represents the current output current. Thus, the measured values of the chopping parameters can be correlated with different actual parameters depending on the current state of the clock control. Finally but not least, when measuring the chopping parameters, the current barrier is particularly advantageously permeable for the measurement.

[0015] In the prior art of such power converters, almost without exception, each measurement first leads through a (sometimes small, i.e., "fast") measurement low-pass filter, the series resistance of which is relatively low-ohmic and the filter capacitor of which is relatively small. Apart from the above, the main reason for this is that the clock-controlled power converter is a strong source of electromagnetic interference beams. Such high-frequency interference (the frequency of which is x times the clock frequency, where x >> 1) can not only distort the important average values since it occurs synchronously with the clock as well, but mainly also destroy the information about the time segments, which should actually be contained in some of the measurement signals. This can lead to jitter and fork up to instability.

[0016] For example, in EP 2 446 708 A1, a chopped and possibly disturbed voltage is to be measured, as shown in FIG. 5 there. In contrast, the measures are shown in FIG. 8 there: A measurement voltage divider consisting of two resistors (which are connected in series between the voltage to be measured) reduces the original voltage, which is too high for the evaluation circuit, to a manageable value, which corresponds proportionally to the total resistance of the entire voltage divider to the resistor coupled in parallel with the evaluation circuit. In order to low-pass filter this reduced measurement signal, a small capacitor is connected in parallel with this parallel-coupled resistor, and the two components are connected to the circuit ground. Although the measurement circuit there is itself the same as the simplest variant of the current measurement device proposed for the clock-controlled power converter, it is significantly different: No current to be measured flows through the measurement circuit there, especially not through the capacitor involved in the measurement low-pass filter, since a voltage is to be measured. The two resistors there are usually extremely high-ohmic, as for the measurement voltage divider, and the capacitor is very small, e.g., in the nanofarad range.

[0017] For example, current measurements with filtering are shown in WO 2008 / 132501 A2. A similar measurement structure, namely a measuring voltage divider, is shown in the upper left of FIG. 9 there, which measuring voltage divider has a capacitor connected in parallel only with one of the voltage divider resistors. Differently from the above, the actual current measuring resistor is connected in parallel therewith, which actual current measuring resistor has a significantly lower resistance than the two voltage divider resistors, i.e., has a significantly higher conductance compared to them. However, since the capacitor mentioned is connected in parallel with the resistor facing away from the circuit ground, there is a measuring high pass. In addition, the evaluation circuit provided therefor must be able to process negative measurement signals.

[0018] In order to keep the losses low, the measuring resistor for current measurement must be as low-resistance as possible. Values from 0.1 ohm to 0.5 ohm, preferably up to 0.25 ohm, for a concentrated current measuring resistor have proven useful for a switching power supply or a power supply device or an LED operating device with a nominal output power of 40 W. Each filter connected thereto is thus again significantly higher-resistance. A common method is to make each filter downstream of the actual measurement, such as the measuring low-pass filter already mentioned, as high-resistance as possible, or if the measurement itself already operates at a very high resistance, as is the case, for example, in voltage measurement, then to integrate it into the measurement.

[0019] However, if the measurement voltage across the low-resistance current measuring resistor does not meet the requirements for resolution, it can be linearly amplified. Since most linear amplifiers have negative feedback, i.e., include, for example, an operational amplifier and a compensation branch, the filter connected upstream of this amplification can be realized by negative feedback or by the compensation branch. Good linear amplifiers usually operate in an inverting manner, which either requires subtracting the measurement signal from a known constant parameter or requires a negative measurement signal. For a negative measurement signal, the amplifier is bipolar-powered, but this is not usually the case. Non-inverting amplifiers are satisfied with a simple voltage power supply but encounter drift and offset problems. The measures required to eliminate these inaccuracies are complex and expensive.

[0020] The losses in all current measuring resistors are based on the effective value of the parameter present thereon. The worse the continuity of the parameter present, the higher these effective values are compared to the average value. Thus, the actual losses are not only proportional to the average value of the parameter to be measured, but the more discontinuous the parameter to be measured, the greater the actual losses. The losses in the current measuring resistor are proportional to the square of the effective value of the current to be measured. The effective value, which well reflects the degree of discontinuity, is furthermore always stated as the ratio to the time average value of the same current. The square of this ratio corresponds to the energy-saving factor (Einsparfaktor).

[0021] As shown above, in a clock-controlled electronic power converter, a reasonable measurement of the chopping parameter results in particularly high measurement losses, especially when the parameter to be measured is the chopping current. In the applications provided for this purpose, the effective value can be up to three times as large as the relevant average value, and in the case of actual research, it is 2.36 times as large, and the measurement losses are correspondingly large.

[0022] Finally but not least, if the average value of the same chopping current is mainly to be evaluated there, then a measurement that is purely proportional to the time curve or the instantaneous value of the chopping current is particularly disadvantageous for the resolution in the evaluation circuit. Summary of the Invention

[0023] The technical problem to be solved by the present invention is to provide a current measurement device that generates lower losses for the evaluation circuit connected thereto than known solutions at a given resolution, or does not generate higher losses for a higher resolution, but at most generates the same losses as those achieved with a smaller resolution according to known solutions, or is able to achieve a combination thereof, i.e., reduced losses while the resolution is increased. For example, a known solution is a centralized current measurement resistor through which most of the chopping current flows. In addition, the technical problem to be solved by the present invention is to provide a current measurement device for a clock-controlled electronic power converter that can be connected in series in any branch of the power converter topology through which a current with a high alternating current component flows.

[0024] In addition, the sought-after current measurement device should avoid negative measurement signals even when the parameter to be measured is partially negative.

[0025] Finally, it may be advantageous if at least part of the logarithmic transfer characteristic curve required for the entire system including the power converter, its measurement, its regulation, and its load has already been implemented in the current measurement device. In particular, it is sought to separately weight and evaluate the time average value of the chopping current to be measured (e.g., in a controlling microcontroller), and in parallel, to separately evaluate and weight the chopping current (i.e., the measurement signal of the instantaneous value without significant low-pass filtering).

[0026] In addition, the technical problem to be solved by the present invention is to provide a regulating circuit designed for such a current measurement device that can evaluate the signal of the current measurement device.

[0027] According to the present invention, the above technical problems are solved with respect to the current measurement device by using the features according to the present invention, and with respect to the regulating circuit by using the features according to the present invention.

[0028] The present invention extends the central concept of the present invention in a particularly advantageous manner.

[0029] Different from the common method of designing all measurement filters to be as highly resistive as possible and connecting them in parallel with the actual measurement, it is proposed here to place the filter in the power path. Thereby, it is possible that the alternating current component of the current to be measured can flow through the so-called average capacitor with almost no loss. The current measuring device can simultaneously form a plurality of measurement voltages, which can be separately evaluated by an evaluation circuit in the functional block, wherein the functional block is designed to apply the current measuring device and is a component of the control circuit of the clock-controlled power converter.

[0030] Therefore, the filter for current measurement is placed in the power path. Thereby, the filter is very low resistive, i.e., contrary to the common method of designing each filter, such as the aforementioned measurement low-pass filter, to be as highly resistive as possible for measurement purposes. Therefore, the current measuring device for the clock-controlled power converter includes at least two independent sensors, i.e., for example, two different resistors or two pure ohmic branches (which are connected in series with each other and thus together contact three nodes), i.e., at least two independent current measurement resistors, and the current to be measured flows uniformly at least partially through these current measurement resistors. The node contacted by the two current measurement resistors or by the two pure ohmic branches is the second node. The current to be measured completely flows through at least one of the current measurement resistors (which is also referred to as the "series current measurement resistor"), and only partially flows through at least one of the remaining current measurement resistors. Since a filter capacitor that completes the required filter and thus the current measuring device for the clock-controlled power converter is connected in parallel therewith, at least one of the remaining current measurement resistors will also be referred to as the "parallel current measurement resistor" in the future. The filter capacitor connected in parallel therewith is the average capacitor mentioned at the beginning.

[0031] For reasons of tolerance, noise, or performance, the series current measurement resistor and the parallel current measurement resistor can each be composed of a plurality of individual resistors, which can be connected in series or in parallel with each other, or in any combination thereof. The current measuring device for the clock-controlled power converter includes three nodes, i.e., the third node into which the current to be measured flows, the first node from which the current to be measured flows and which is usually connected to the circuit ground, and the second node connected to the two independent current measurement resistors. There is a pure ohmic branch between the first and second nodes, and the value of this pure ohmic branch corresponds to the first current measurement resistor. There is again a pure ohmic branch between the second and third nodes, and its value corresponds to the second current measurement resistor.

[0032] The first end of the mean capacitor is additionally connected to the second node. The second end of the mean capacitor is either connected to the first node or to the third node. The total value of all its individual resistors that are directly or indirectly connected to the second node and at the same time directly or indirectly connected in parallel with the mean capacitor corresponds to the shunt current measuring resistor, and the total value of all the remaining individual resistors of the current measuring device under consideration corresponds to the series current measuring resistor.

[0033] For the same reason, the mean capacitor can also consist of a parallel circuit of a plurality of individual capacitors, which can also have different structural types, which shows another reason for it. For example, an electrolytic capacitor and a ceramic capacitor can be connected in parallel, or a thin film capacitor and a ceramic capacitor can be connected in parallel: the latter provides speed and the former provides the necessary capacitance. Since the maximum voltage occurring there is below 5 V, a series circuit of capacitors is excluded.

[0034] Thus, the current measuring device for a clock-controlled power converter comprises two sensors connected in series independently of one another, the conductance of the series current measuring resistor and the admittance generated by a parallel circuit formed by the mean capacitor and the shunt current measuring resistor.

[0035] When the second end of the mean capacitor is connected to the first node, the admittance is between the first and second nodes, so the first current measuring resistor has the function of the shunt current measuring resistor, and the second current measuring resistor has the function of the series current measuring resistor or the function of the conductance. However, if the second end is connected to the third node, then the functions are interchanged, so that the first current measuring resistor forms the series current measuring resistor or the conductance, while the second current measuring resistor provides a share for the admittance as the shunt current measuring resistor, the susceptance of which is formed by the mean capacitor.

[0036] In this current measuring device, it is fully utilized that the averaging capacitor guides most of the AC current component of the current to be measured without loss and is charged to a voltage which starts from a smaller value and approaches the actual time average of the current to be measured and will be referred to as the "measured average" in the future. Exactly, this measured average corresponds to the low-pass filtered instantaneous value of the current to be measured with a time constant which corresponds to the time constant of the admittance and will be determined more precisely below. In addition to approaching the actual time average of the current to be measured from below, the measured average also always has a ripple at the clock frequency of the power converter under consideration. From this it follows that the measured average can sometimes also be higher than the actual time average of the current to be measured. However, the theoretical center line of the measured average is always below the actual average, unless the averaging capacitor is infinitely large. All this also applies if the current to be measured is partly negative. As long as its measured average is positive, which is almost always the case when active power is transferred from the input to the output of the power converter, negative measurement signals are avoided by this averaging formation. Instead, the sign of the measured average indicates the direction of the power flow. In addition to separating the current measuring resistor by the second node, the averaging capacitor, which has recently been placed in the power path, is also the central element of the current measuring device for a clock-controlled electronic power converter, as can also be read from the name of the two current measuring resistors connected to it. The current to be measured is advantageously the operating current of an actively controllable power transistor within the electronic power converter under consideration, or generally the current in a branch of the power converter topology, which is chopped or can be partly negative and thus contains a high AC current component, since the electronic power converter under consideration operates in a clock-controlled manner. According to the invention, the effective value of the current to be measured is at least 1.4 times its time average. The provided current measuring device is precisely to determine this, and therefore it must also exist. This is reliably provided in a current whose effective value corresponds at most to 8 times its time average. The higher this relative effective value, the more precisely the provided current measuring device operates, even when its actual purpose (which is also to determine the time average of the current) may be lost in the process. Alternatively, the DC current error can be determined in a practically pure AC current, and in the case of a ground point shift, the exact time curve of the pure AC current component can be measured.

[0037] Another advantage of the current measuring device for a power converter is the possibility to simultaneously provide two measurement signals which are related to each other but still different.

[0038] If the mean-value capacitor is connected to the first node, then the measured mean value from the current measuring device for the clock-controlled power converter can be tapped directly as a first measurement voltage between the second and the first node in parallel with its admittance. The sum of the measured mean value and the measured value for the instantaneous value of the current to be measured can be tapped as a third measurement voltage between the third and the first node of this arrangement.

[0039] However, if the mean-value capacitor is connected to the third node, then the measurement voltage for the measured mean value again exists between the third and the second node in parallel with the admittance, that is, correspondingly. The measured value for the instantaneous value of the current to be measured can be tapped as a second measurement voltage across the conductance between the second and the first node of this arrangement. The sum of the measured mean value and the measured value for the instantaneous value of the current to be measured can again be tapped as a third measurement voltage between the third and the first node.

[0040] Accordingly, the third node is also referred to as the summing point, and the first node can be connected to the circuit ground. The usually sought peak value of the current to be measured can be identified either on the measured value for the instantaneous value or on the summed measurement voltage.

[0041] As has been the case thus far, in the following description, these currents, which flow into the functional block for the purpose of identifying the measurement voltage and generally flow out again in the direction of its circuit ground and are small (e.g., one-thousandth or less) relative to the current to be measured, are not considered further in principle. The functional block is arranged downstream of the disclosed current measuring device for the clock-controlled power converter and belongs to the regulating circuit for the same power converter.

[0042] The time constant of the admittance should advantageously result in twice the above-mentioned corner frequency of the entire power converter, so that a significant mean-value change of the current to be measured can still be correctly mapped by the measured mean value, that is, by the value low-pass filtered from its instantaneous values. The fluctuations in the first measurement voltage across the parallel circuit consisting of the mean-value capacitor and the parallel-connected current measuring resistor, that is, the fluctuations in the measured mean value in parallel with the admittance, should be attenuated by less than 10% relative to the actual fluctuations in the time mean value of the current to be measured at twice this corner frequency.

[0043] The time constant of the admittance is preferably greater than 0.2 times the maximum occurring cycle duration of the clock control of the power converter, and the time constant is at most 20 times this cycle duration. Particularly preferably, the time constant of the admittance is greater than 0.6 times the maximum occurring cycle duration of the clock control of the power converter, and the time constant is at most 5 times this cycle duration. Due to the relatively small value of the shunt current measuring resistor, the averaging capacitor can for this purpose have a relatively large capacitance, which is advantageous for guiding the AC current component of the current to be measured.

[0044] Conductance is not in principle considered for calculating the time constant of the admittance, because the current flowing through the series-connected current measuring resistor, i.e., the current to be measured, still always at least partially flows through at least one storage inductor of the power converter topology, and this storage inductor, due to its significantly higher impedance, respectively applies this current or current component. In other words, when a very large impedance is also connected in parallel with the shunt current measuring resistor, the value of this shunt current measuring resistor remains approximately constant.

[0045] Basically, only the DC component or the average value of the current to be measured flows through the shunt current measuring resistor, and losses corresponding to this average value are generated there, which are significantly lower than the losses corresponding to the effective value of the current to be measured, for example, with an energy-saving factor of 1.96 to 64 lower. Because the losses increase with the square of the effective value of the current to be measured. The ratio of the effective value of the current to be measured to the time average value is at least 1.4, from which it can be said that there is a notable AC current component in the current to be measured. Using (1.4) 2 = 1.96, the above-mentioned minimum energy-saving factor for loss reduction is produced. In the actually investigated case, the ratio between the effective value and the average value is 2.36, and its square results in an energy-saving factor of 5.57 for possible loss reduction. Conversely, all this also means that there must be a notable DC component in the current to be measured, because as already mentioned at the beginning, this DC component usually even represents the most important parameter for evaluation. Therefore, a current measuring device for a clock-controlled power converter is not suitable for measuring a pure AC current, unless a measuring rectifier is connected upstream of it, the current to be measured flows through the input of this measuring rectifier, and the current measuring device is connected to the output of this measuring rectifier. The current measuring device is then suitable for determining the "rectified average value" with low losses. For a current measuring device without a measuring rectifier, the effective value of the current to be measured should not exceed 8 times its average value in order to still form a sufficient DC component.

[0046] Since the AC current component of the current to be measured is conducted losslessly through the susceptance of the averaging capacitor, the current measuring resistor connected in parallel with the averaging capacitor can be implemented with a higher resistance than the current measuring resistor connected in series with both, i.e., higher resistance than the series-connected current measuring resistor, and it can even be higher resistance than the single concentrated current measuring resistor in known solutions. This enables a higher average measurement voltage and thus a higher resolution in the functional block. The value of the parallel-connected current measuring resistor can be slightly less than the energy-saving factor higher than the value of the concentrated current measuring resistor of the prior art, for example, a factor of 5 higher in the case under study, compared to which a five-fold finer resolution of this measurement in the functional block can be achieved. For the series-connected current measuring resistor, a remainder of at most the energy-saving factor, i.e., 0.57, is required. This resistor must therefore approximately have half the value of the single concentrated current measuring resistor in the prior art, or its conductance is approximately twice as large as the conductance of a concentrated current measuring resistor.

[0047] Because the series-connected current measuring resistor or conductance through which the current to be measured fully flows provides information about the time segments. Since the steepness to reach a given level is almost always irrelevant (differentiated evaluation is very rare), it can also be smaller than in known solutions with a concentrated current measuring resistor. This enables a series-connected current measuring resistor that can clearly have a lower resistance than a single concentrated current measuring resistor. Although the loss of this series-connected current measuring resistor is based on the effective value of the current to be measured, it generates less loss than in known solutions with a concentrated current measuring resistor according to the small value of this resistor.

[0048] The degree of chopping, i.e., the ripple of the measurement signal required at least to reliably identify and evaluate the time segments in the functional block, determines the value of the series-connected current measuring resistor. Its loss should be less than the loss allowed for the entire measurement circuit. The parallel-connected current measuring resistor connected in parallel with the averaging capacitor uses the difference therein, and the conductance value of this parallel-connected current measuring resistor is obtained by dividing the square of the maximum occurring measured average value taken for averaging by this remaining loss. The smaller the remaining loss, the larger this conductance value, and thus the smaller the average measurement voltage, and thus the smaller the possible resolution in the functional block. The conflicting parameters of "safety of time segment identification" and "resolution of the average value" must be weighed against each other.

[0049] Here it is advantageous that the active conductance value of the admittance is related to the nominal output power Pnom of the clock-controlled electronic power converter and corresponds to 40V 2 / Pnom and 1000V 2The resistance value between / Pnom. The active conductance value of the admittance preferably corresponds to the following resistance value, which is related to the nominal output power Pnom of the clock-controlled power converter and generates 80V 2 / Pnom and 400V 2 The ohmic value between / Pnom and 400V. Particularly preferably, the active conductance value of the admittance corresponds to the following resistance value, which is related to the nominal output power Pnom of the clock-controlled power converter and generates 80V 2 / Pnom and 400V 2 The ohmic value between / Pnom and 400V, and the capacitance of the averaging capacitor is also related to the nominal output power Pnom and is from Pnom*25ns / V 2 to Pnom*500ns / V 2 .

[0050] In one embodiment, here, the total conductance value of the conductance is 2 to 100 times higher, especially 5 to 40 times higher, and particularly advantageously 10 to 25 times higher than the active conductance value of the admittance.

[0051] The more precisely the possible shape and value range of the current to be measured and the purpose of the measured signal are known, the further the minimization of the measurement loss can be advanced, which is particularly included in the following further description and drawings. It is shown that the conductance can be 5 to 20 times higher than the conductance value of the shunt current measuring resistor.

[0052] Both, i.e. the shape and the purpose, are closely related to the set power converter topology. Therefore, two of them are representatively considered below as examples of current measuring devices for clock-controlled power converters: an isolated buck-boost converter or flyback converter and a half-bridge. In the case of the half-bridge, time segmentation is less important than in the case of the flyback converter, which usually should also operate as a power factor corrector at the same time. Since just this kind of circuit almost always operates in the so-called "critical conduction mode (CRM)" or "transient conduction mode (TCM)" or "valley detect mode", the current to be measured is the current flowing through the power transistor, which is basically triangular during the conduction time. All important peaks of this current must be reliably identified once during the duration of each cycle of the clock control of this power converter, and then the power transistor must be turned off each time. In contrast, in the case of the half-bridge, the information about time segmentation is mainly used for overcurrent identification of at least one of the power transistors, so the entire half-bridge should be briefly turned off or at least be pulled back. Such an overcurrent can occur, for example, in the case of a short circuit on the output side, when the grid voltage is switched on or in the case of a voltage pulse on the grid line. Due to the internal inertia of the half-bridge, which mostly operates resonantly and thus in a "continuous conduction mode (CCM)", it does not matter whether the overcurrent shutdown occurs immediately or only after half a cycle duration. Here, "continuous" should not be understood in the sense of uniform or "having a high DC current component", but in the sense of "non-interrupted".

[0053] Since the peak value of the current to be measured in the half-bridge is always at least of the same order of magnitude as the relevant average value, the measured average value can be transferred uncorrected to the functional block as the first measurement voltage, and the measured value for the instantaneous value added to it can be considered independently as the third measurement voltage for all types of safety shutdown. For this purpose, the second end of the average value capacitor is advantageously connected to the first node.

[0054] The actual time average of the current to be measured is always higher in the average than the measured average that can be taken as the first measured voltage at the second node (unless the admittance has an infinite time constant, but this has already had to be excluded above), and the measured average corresponds to the low-pass filtered signal of the instantaneous value of the same current. To obtain the actual and higher average value from this value, the time average of the measured signal of the instantaneous value on the conductance must be added to it, and the sum must be low-pass filtered or averaged again. Thus, the actual average value corresponds to the sum formed by the measured average value and the measured instantaneous value, and these two values are averaged together again. Its circuit-technical equivalent (which also further improves the signal-to-noise ratio in the functional block due to the higher absolute measured value) is to measure only the third measured voltage at the summing point or the third node, and this third measured voltage is passed unchanged or through a so-called "fast sequence filter" to the second input of the second functional block for the safety device, and in parallel to the first input of the first functional block for current regulation through a very slow measuring low-pass filter (which is subsequently also referred to as the "slow sequence filter"). To keep the current flowing through such sequence filters and ultimately into the functional block negligible, the slow sequence filter as well as the fast sequence filter can again be classically dimensioned as high-impedance.

[0055] Without passing the third measured voltage in parallel through the slow sequence filter, a simpler evaluation circuit (especially in the form of a prefabricated power converter - controller integrated circuit) can also be coupled to the provided current measuring device. These controller integrated circuits almost always have only one input for the measured value of the current to be measured (which is usually the power transistor current). Therefore, this one input is particularly advantageously connected to the third node or to the summing point. Because such controller integrated circuits require information about time segments and information about the average value. However, since in the current measuring device for a clock-controlled power converter, the scales of the average value and the instantaneous value are usually different, especially the scale of the average value is increased compared to the scale of the instantaneous value, starting from the first measured voltage at the low-impedance second node, that is, from the hitherto unused measured average value, at least one rated value for the controller integrated circuit must be correctedly influenced. If this rated value is not led out through a connection on the controller integrated circuit or preset from the outside, but is formed by an internal reference as usual, then the application of the provided current measuring device in combination with such a simple prefabricated controller integrated circuit for a clock-controlled power converter is excluded. Therefore, this possibility is not further considered.

[0056] Measuring a low-pass filter or a normal sequence filter is also meaningful for separately evaluating two signals on a current measuring device for a clock-controlled power converter, and is particularly advantageously applied, for example, to a half-bridge. Here, a first measured voltage, i.e., the measured average value, can be passed through a slow sequence filter in the form of a low-pass filter to a first input of a first functional block, which further smoothes the first measured voltage with a time constant that at least corresponds to the time constant of the admittance, i.e., the time constant of a parallel circuit composed of an average value capacitor and a parallel-connected current measuring resistor, or can be up to 50 times or even up to 500 times larger than the time constant of the admittance. Alternatively, the slow sequence filter is dimensioned such that at its output, all voltage fluctuations with a frequency higher than 20 kHz are suppressed by at least 90%. And the measured value of the instantaneous value of the power transistor current added to the measured average value, i.e., the third measured voltage (from which at least one peak can be determined), is passed through a fast sequence filter to a second input of a second functional block. The fast sequence filter can be a direct connection or a simple series resistor for the purpose of impedance matching with the functional block, which together with the input capacitance of the second input acts as a very fast low-pass filter. However, it can also be a complete fast low-pass filter. Alternatively, the fast sequence filter can be a band-pass filter in order to particularly preferably forward typical error signals to the functional block, or it can finally be a band-stop filter in order to prevent false triggering of a safety shutdown. The output of the slow sequence filter can in turn be connected to the first input of the first functional block provided therefor via a series resistor.

[0057] When applying a current measuring device for a clock-controlled power converter in a flyback topology, or more generally, whenever a power converter should also operate as a power factor corrector while performing its power conversion and transfer, information about time segments becomes so important that not only must any sequence filter be adjusted, but it is even particularly advantageous if the functional sequences consisting of admittance and conductance can be interchanged in the current measuring device. Because it is more advantageous to provide a pure instantaneous value signal instead of a pure average value signal (in the form of the measured average value), which can be taken as a measured value for the instantaneous value of the current to be measured in the form of a second measuring voltage in parallel with the conductance of the current measuring device at a second node. The second measuring voltage also contains the usually sought peak value. For this purpose, the second end of the average value capacitor is connected to a third node. The series-connected current measuring resistor or conductance is formed by a first current measuring resistor, and its first end is particularly advantageously located on the circuit ground or connected to the first node, and its second end marks the second node shared with the parallel-connected current measuring resistor (formed by a second current measuring resistor) and the average value capacitor. As in the case of the half-bridge above, the sum of the measured value of the average value and the measured value of the instantaneous value can be taken as a third measuring voltage at the summing point or the third node, which is connected, for example, to the reference electrode of a power transistor whose operating current is to be measured.

[0058] Therefore, the slow sequence filter for the measured value of the average value is connected here to the summing point, i.e., the third node, and the fast sequence filter is connected here to the second node between the conductance and the admittance. Here, the two filters can be implemented as above for the half-bridge. Here, a band-pass filter is beneficial for correct time control, while a band-stop filter avoids control errors such as jitter. Since only the time segments are evaluated at the second input of the second functional block provided for this purpose, the reduced amplitude of the peak signal determined by the instantaneous value measurement does not cause any defects compared to the known solutions with a concentrated current measuring resistor.

[0059] The technical problems related to the regulation circuit are solved by designing a regulation circuit for the current measuring device, which has:

[0060] - A first functional block for implementing current regulation of a regulated power supply, which uses the first or third measuring voltage as an input signal representing the current at present,

[0061] - A second functional block for implementing overcurrent shutdown of a regulated power supply, which uses the second or third measuring voltage as an input signal representing the current at present,

[0062] It also has an adaptation network, which may include first and / or second sequence filters, and the adaptation network is connected between the measured voltage and the functional block. By filtering the two measured voltages differently, advantages in the evaluation in the regulating circuit are advantageously produced, so that it can operate very effectively.

[0063] In a preferred embodiment, the first sequence filter is a low-pass filter and has a time constant that is 0.01 to 100 times the time constant of the admittance. Through this design, sufficient smoothing of the signal is advantageously achieved.

[0064] In one embodiment, the second sequence filter consists of a direct connection or a series resistor between the current measuring device and the second input of the second functional block, without a connection to the reference potential. This advantageously ensures a very simple, inexpensive and effective measurement of the instantaneous value.

[0065] In a further embodiment, the second sequence filter is a low-pass filter having a time constant between 10 ns and 100 μs, particularly advantageously between 100 ns and 10 μs, and the second sequence filter has a connection to the reference potential. The smoothing of the low-pass filter produces a more continuous measurement signal, which can be evaluated more simply.

[0066] In a further embodiment, the second sequence filter consists of a band-pass filter in order to advantageously forward typical error signals preferably to the functional block or to facilitate correct timing control. In a further embodiment, the second sequence filter consists of a band-stop filter in order to advantageously prevent false triggering of a safety shutdown or to prevent jitter.

[0067] Preferred embodiments can be found in the present invention and the entire disclosure, wherein in the drawings, there is not always a detailed distinction between the device and the usage aspects; implicitly in any case, the present disclosure should be read in view of all categories of the present invention.

[0068] Further advantageous extensions and design options for the current measuring device of a clock-controlled electronic power converter are given by the present invention and the following description. Description of the Drawings

[0069] Other advantages, features and details of the present invention are given according to the subsequent description of the embodiments and according to the drawings, in which the same or functionally identical elements are provided with the same reference numerals. The embodiments and the drawings have exemplary characteristics only, and thus do not limit the scope of protection to their direct content. In these drawings:

[0070] Figure 1shows the basic structure of a regulated electronic device, which includes a clock-controlled power converter having a regulation circuit that in turn includes two functional blocks,

[0071] Figures 2a to 2f shows different possible implementations of a current measuring device for a clock-controlled power converter,

[0072] Figure 3a shows a comparison between the instantaneous value, the average value, the squared value, and the square of the average value of the current to be measured in a bridge leg that is a resonant load of a clock-controlled power converter,

[0073] Figure 3b shows a comparison between the instantaneous value, the average value, the squared value, and the square of the average value of the current to be measured in a buck converter operating in synchronous operation as a clock-controlled power converter,

[0074] Figure 3c shows a comparison between the instantaneous value, the average value, the squared value, and the square of the average value of the current to be measured in a boost converter or flyback converter for power factor correction operating as a clock-controlled power converter,

[0075] Figure 3d shows a comparison between the instantaneous value, the average value, the squared value, and the square of the average value of the current to be measured in a buck converter operating with continuous current as a clock-controlled power converter,

[0076] Figures 4a to 4h shows different possible configurations of the admittance and conductance of a sequence filter and a functional block. Detailed Description

[0077] Figure 1The basic structure of a regulated power supply device 700 is shown, which can equally be a plug-in power supply, a switching power supply, a laboratory power supply, a DC current converter, a DC voltage converter, an active current source, an active voltage source or an operating device for a light source, preferably for a light-emitting diode, as the simplest power supply device. The power supply device includes a power component 600 having at least one clock-controlled electronic power converter, the electronic power converter comprising at least one storage capacitor and at least one storage inductor or storage coil or storage choke and at least one rectifier diode and at least one actively controllable power transistor. For this purpose, the three simplest power converter topologies are the buck converter, the boost converter and the buck-boost converter (Buck-Boost or Flyback). The buck-boost converter requires two storage capacitors. Alternatively, if two storage inductors that can also be coupled are provided and the storage capacitor is moved again between the power transistor and the rectifier diode, then a converter can be constructed therefrom. The Zeta converter and the SEPIC (single-ended primary inductor converter), i.e., the Zeta converter at its input and the SEPIC at its output, also require a second storage capacitor starting from the converter. If two actively controllable power transistors are used in the same power converter topology, then in particular a so-called half-bridge converter is formed, which in addition to at least one additional rectifier diode also requires at least one further storage capacitor connected in series with the storage inductor, and this further storage capacitor absorbs the DC voltage component of the half-bridge output voltage. If this series capacitor and the storage inductor (which is advantageously at least partially implemented by a transformer in the case of a half-bridge) together form an oscillating circuit having a natural frequency just below the lowest clock frequency, then a resonant LLC converter with ZVS switch release is formed. If a pure resonant capacitor is additionally connected in parallel with the transformer, then a so-called resonant LCC converter is produced in the case of also regulating its natural frequency as above with ZVS switch release. This means that each conduction process of the power transistors participating in the clock-controlled power converter is realized without voltage, thereby avoiding any conduction losses. In addition, the down-frequency effect of the purely ohmic part of each load can be used to bring all components connected to the output of the LLC or LCC converter below the clock frequency.

[0078] Furthermore, the power component 600 includes grid input terminals 603, 604 for inputting electrical energy, a radio interference suppression filter, a grid rectifier, and almost always a power factor correction stage following it, which can also be configured as a charge pump and is supplied by at least one of the above-mentioned power converter topologies (which operates as an actual clock-controlled electronic power converter). Finally, the power component 600 also includes an output terminal for the load, at which the output voltage and the load current can be sent, which in some cases also returns from the load together with a measurement input to a power component such as of the power component. For the sake of completeness, all these details are mentioned, except for exceptions that are not part of the present invention and are therefore not shown within the power component 600.

[0079] The current measuring device 100 is connected in series with a branch of the power converter topology within the power component 600, in which the current 20 to be measured having a DC current component and simultaneously a high AC component flows, such that its effective value is at least 1.4 times and at most 8 times as large as its average value. The current measuring device includes a node 1 (which is usually connected to the circuit ground and from which the current 20 flows out), an intermediate node 2 at which a first measurement voltage 21 or a second measurement voltage 22 can be taken, and a third node 3, also referred to as a summing point, at which the current 20 flows into the current measuring device and a third measurement voltage 23 can be taken.

[0080] The regulated power supply device 700 includes a regulation circuit 500 as a second important area, and the regulation circuit 500 can in turn be composed of at least components: a first functional block 205, a second functional block 206, and a clock generator 300. At least one clock generator 300 generates at least one signal 301, and at least one actively controllable power transistor of the clock-controlled control power converter in the power component 600 is controlled by using this signal. If the signal 301 must be amplified or otherwise processed before finally controlling the power transistor, the circuit required for this belongs to the power component 600. The signal 301 generally includes at least two channels, namely one channel for power factor correction and at least one other channel for the actual clock-controlled power converter. The regulation circuit generally has a direct input terminal 203, and the entire power supply device can be externally controlled through this input terminal. An interface (not shown) for the wired or wireless communication protocol from DALI to Thread, that is, a 1-10V dimmer, such as a DIP switch for regulating the load current, etc., can be connected to this input terminal 203. However, here, the regulation circuit particularly additionally includes a second input terminal 212 on the second functional block 206 and an adaptation network 150, and this adaptation network is connected between the second and first input terminals 211 of the first functional block 205 and the measurement voltages 21, 22, 23 from the current measurement device 100.

[0081] Another signal, such as a signal 601 for the output voltage and / or a signal 602 for the input voltage or for the temperature of a critical component, can be directly transmitted from the power component 600 to the first functional block 205. This component is temperature-sensitive, such as an electrolytic capacitor, or generates a large amount of heat, such as an inductor, an active power transistor, or a rectifier diode.

[0082] As described above, the first functional block 205 is coupled to the first input terminal 211. The second functional block 206 is coupled to the second input terminal 212. The first functional block implements current regulation for the regulated power supply device 700. For this purpose, the filtered signal is input into the first functional block at the input terminal 211, and this filtered signal represents the first measurement voltage 21 or the third measurement voltage 23. As described above, the measurement voltages 21, 23 are low-pass filtered and therefore represent the average value of the current to be measured in the power component 600. This signal is used to regulate the current flowing through the current measurement device 100. For this purpose, the first functional block 205 has a current regulation, which is known per se and is therefore not described in detail here. The first functional block sends an output signal to the clock generator 300, and this output signal represents the conduction time and duration of at least one power transistor in the power component 600. Thus, the clock generator 300 generates a control signal 301 for at least one power transistor in the power component 600.

[0083] The second functional block implements a fast overcurrent shutdown in order to quickly shut down in the event of an unforeseen current and thus protect at least one power transistor. To this end, the second functional block 206 has a second input 212, and there a signal representing the second measurement voltage 22 or 23 is input, which, as described above, represents the instantaneous value of the measured current. If this instantaneous value rises above a predefined threshold, then the second functional block 206 sends a signal to the clock generator 300, which quickly shuts down the power transistor in order to thereby end the overcurrent situation. Since the second functional block must react quickly to avoid damaging the transistor, the signal input into the second input 212 is either not filtered at all or the filter contained in the adaptation network 150 is very fast in order to only filter out unwanted short voltage peaks.

[0084] Figure 2a A first basic structure of a current measuring device for a clock-controlled power converter is shown. Node 1 and 2 are connected via a first current measuring resistor 11, and a second current measuring resistor 12 is present between node 2 and 3. The third measurement voltage 23 can be measured at node 3 or at the summing point 3, and the first measurement voltage 21 is measured at node 2, which first measurement voltage corresponds to the above-defined "measured average value" of the current 20 to be measured. Since the averaging capacitor 10 is connected between node 1 and 2 in parallel with the first current measuring resistor 11, an admittance is generated from this parallel circuit and a conductance is generated from the second current measuring resistor 12. Thus, here, the second current measuring resistor 12 is a series current measuring resistor, and the first current measuring resistor 11 forms a parallel current measuring resistor.

[0085] The second measurement voltage 22, which is proportional to the instantaneous value of the current 20, is in parallel with the second current measuring resistor 12 and can therefore not be measured in a simple manner here, since it floats around the measured average value 21 when viewed from node 1.

[0086] Figure 2b A variant that is obvious based on foreseeable dimensions is shown, in which, in a current measuring device for a clock-controlled power converter, a plurality of single resistors with similar values are used for each of the first and second current measuring resistors. The first current measuring resistor (which provides a share of the admittance as a parallel current measuring resistor) consists of a series circuit of the resistors 11a and 11b between node 1 and 2, and the second current measuring resistor (which, as a series current measuring resistor or as a conductance, can be significantly lower in resistance than the first current measuring resistor as described above) consists, for example, of a parallel circuit of three single resistors 12a... 12c.

[0087] The measurable measurement voltages 21 and 23 correspond in shape and position to Figure 2aMeasured voltage.

[0088] If the current measuring device according to Figure 2a or Figure 2b becomes overall low-resistance, then a circuit as shown in Figure 2c is possible. Its averaging capacitor here consists of at least two single capacitors 10a and 10b connected in parallel to form a susceptance for admittance, and the first current measuring resistor connected in parallel therewith in turn consists of a parallel circuit of two series circuits 11a + 11b and 11c + 11d of single resistors. The two lines located in the middle respectively can be connected to each other (not shown) so as to change from the parallel circuit of the series circuits into an equivalent series circuit of parallel circuits. The conductance or the series current measuring resistor or the second current measuring resistor between nodes 2 and 3 here consists of a direct parallel circuit of, for example, four single resistors 12a... 12d or more single resistors. The low voltage on each averaging capacitor 10a and 10b makes the replacement series circuit of single capacitors meaningless.

[0089] The measurable measuring voltages 21 and 23 correspond in shape and position to Figure 2a the measured voltage.

[0090] Figure 2d Fig. shows a second basic structure of a current measuring device for a clock-controlled power converter. Node 1 and 2 are connected by a first current measuring resistor 11, and a second current measuring resistor 12 exists between nodes 2 and 3. As described above, a third measuring voltage 23 can be measured at node 3 or the summing point 3, and, contrary to the above, a second measuring voltage 22 is measured at node 2, which extends in proportion to the instantaneous value of the current 20 to be measured. Here, the averaging capacitor 10 is connected in parallel with the second current measuring resistor 12 between nodes 2 and 3, so that an admittance is generated from this parallel circuit, and as elaborated above, the measured average value 21 can be measured in parallel with this admittance, which measured average value now exists as a measuring voltage between nodes 2 and 3 and thus "floats". Now, the first current measuring resistor 11 forms a conductance that is connected to node 1 and thus provides a "proportional" second measuring voltage 22 that can be measured at node 2. Therefore, here, the first current measuring resistor 11 is a series current measuring resistor, and the second current measuring resistor 12 provides a share for the admittance as a parallel current measuring resistor.

[0091] According to Figure 2dThe lower part of this structure, i.e., from the current 20 to be measured that leaves it through node 2 until the second measurement voltage 22, is known in the prior art. However, in comparison, the conductance of the first current measurement resistor 11 is significantly greater here. Although this conductance results in significantly smaller losses compared to the prior art, it also produces a smaller measurement voltage amplitude of the second measurement voltage 22. As already explained above, it only has to provide information about the time segments, and for this purpose, its smaller amplitude is sufficient. Since the essential novelty lies in the admittance as an additional sensor, from the perspective of the current 20 in series with the lower part:

[0092] The measured average value present thereon, due to the averaging capacitors 10, 10a, 10b still having very low losses, carries the main part of the information about the generally more important average value of the current 20. Since this measured average value is always lower than the actual average value of the current 20 by a value that lies within the range of the instantaneous value of the second measurement voltage 22, it is not important that the measured average value 21 "floats" with the second measurement voltage 22. As the sum of the measured voltage 22 of the instantaneous value and the measured voltage 21 of the measured average value, the third measurement voltage 23 that can be measured at the summing point 3 thus carries the information about the actual average value of the current 20 most precisely.

[0093] Figure 2e shows Figure 2d A variant that is obvious according to the foreseeable dimensional design, in which, in the current measurement device for a clock-controlled power converter, multiple single resistors with similar values are used for each of the first and second current measurement resistors. The first current measurement resistor (which, as a series current measurement resistor, can be significantly lower in resistance than the second current measurement resistor and forms the conductance) consists of a parallel circuit of three single resistors 11a... 11c between nodes 1 and 2, and the second current measurement resistor (which, as a parallel current measurement resistor, provides a share for the admittance) consists of, for example, a series circuit of resistors 12a and 12b between nodes 2 and 3.

[0094] The measurable measurement voltages 22 and 23 correspond in shape and position to Figure 2d the measurement voltages.

[0095] If the current measurement device according to Figure 2d or Figure 2e is to generally become lower in resistance overall, then as Figure 2fThe circuit shown is possible. Its averaging capacitor consists here of at least two single capacitors 10a and 10b connected in parallel to form a susceptance for admittance, and the second current measuring resistor connected in parallel thereto consists of the parallel circuit of two series circuits 12a + 12b and 12c + 12d of single resistors. The conductance or the series current measuring resistor or the first current measuring resistor between nodes 1 and 2 here consists of the direct parallel circuit of, for example, four single resistors 11a... 11d or more single resistors.

[0096] The measurable measuring voltages 22 and 23 correspond in turn in shape and position to Figure 2d the measuring voltages.

[0097] Figure 3a The curve of the current 20 is shown in the line graph above (current I versus time t), which is typical, for example, for a resonant half-bridge as a clock-controlled power converter. Starting from the negative value at the beginning 25 of the conduction phase of the lower power transistor, for example, the current 20 rises rapidly in order to quickly change its sign at the time point 26 and, after a sinusoidal-like curve, is switched off again at the positive end value of the same conduction phase at the time point 28. Usually, shortly before this, at the time point 27, the current 20 exceeds the maximum value. In this form, the lag of the current 20 with respect to the clock frequency of the half-bridge power converter can be clearly seen, which is in principle a segment of the non-rectified and smoothed load current or a part thereof, resulting from the coordination of the rectifier and the load (again with respect to the clock frequency) with a slight inductance in order to achieve ZVS. ZVS means "Zero Voltage Switching" or conduction at zero voltage, whereby each conduction process of at least one clock-controlled power transistor is optimally released.

[0098] The curve 21 of the measured average value of the current 20 is plotted in the same line graph, which has been determined, for example, in the investigated embodiment. The scales of the two curves 20 and 21 are the same, which can be seen in the correctness of the already optically readable average value. The horizontal line 24 corresponds to the actual average value of the current 20. As already explained above and as can be well read here in the comparison of the area parts between the curve 21 and the horizontal line 24, if the measured average value in the clock frequency due to Figures 2a to 2fIf the average value of one of the capacitors 10 in is not infinite and has ripples, then the measured average value 21 (even if it can be larger in some instantaneous time intervals) is always less than the actual average value 24 in the average value. The area part above the measured average value 21 is larger than the unshaded area part below the measured average value. The light shaded rectangular area in the line graph below is needed to prove the desired effect of the present invention.

[0099] Figure 3a The middle line graph also shows the square I of the current 20 on the same scale as the curves 20 and 21. 2 The time curve 400 of the current 20 with respect to time t. Since the value of the loss in the current measuring resistor for determining the current 20 is obtained from the square of the same current according to the known formula Pv = R * I 2 (t) is obtained from the square of the same current. This formula defines the effective value: in a resistor with a specific value R, which constant current results in the same power loss Pv as the actual time-varying current 20 or I(t)? This constant (theoretical) current is the effective value of the actual current I(t).

[0100] However, since we are not interested in this theoretical current at all, but only in the power loss caused by the actual time-varying current I(t) or 20, the above-known formula already describes the solution: the area surrounded by the time curve 40 and the relevant time axis and shown shaded corresponds to the energy that is converted into heat in a resistor with a specific value R by the current 20. If this energy is related to the time spent on it, that is (since it is periodically repeated between the time points 25 and 25', which describe two consecutive conduction processes of the same power transistor of the power converter under consideration), meaningfully, the shaded area between the boundary lines 25 and 25' of the period duration T is divided by this period duration, then an image of the power loss formed in this resistor is produced.

[0101] In Figure 3a The square I of the measured average value 21 of the current 20 is shown in the line graph below 2The time curve 441. If the same resistor with a specific value R is only flowed through by a current corresponding to the actual current I(t) or the measured average value of 20, then here too, the area between the time curve and the associated time axis is represented by shading, and the part between the boundary lines of the same period duration T is an image of the power loss. This current flows through the parallel current measuring resistor and is part of the admittance of the provided current measuring device. The area spanned by the square 441 of this current average value 21 is significantly smaller than the area spanned by the square 400 of the actually to-be-measured current 20, although there are also losses in several time periods in the lower line graph, in which there are zero intervals in the middle line graph.

[0102] A small thought experiment can be used for explanation: Given a constant current of magnitude "2". Since this current is constant, its effective value is also of magnitude "2". However, if the same charge is to be transferred with a duty cycle of 50%, i.e., if rectangular current blocks are to be formed, which are as "wide" as the zero intervals between them, then all the current blocks have a magnitude "4". The square of which is "16". However, since this only applies to half of all time periods, the square of the effective current is "8". This roughly corresponds to the shaded area in the middle line graph. "2" roughly corresponds to the shaded area in the lower line graph, resulting in the area there being significantly smaller than the area in the middle line graph, although there are no zero intervals in the lower line graph.

[0103] The same difference appears in the measured loss when the resistor is used as a single, concentrated current measuring resistor and losses are generated according to the middle line graph, or when the same resistor provides a share for the admittance as a parallel current measuring resistor and only causes smaller losses according to the lower line graph here. The square of the effective value of the current 20 produces a factor compared to its average value, and through the provided current measuring device, the power loss caused by it can be minimized by this factor. Therefore, this factor is also called the "energy-saving factor".

[0104] Figure 3aThese are the measurement results on which the sizing of the invention and of the circuits according to the invention is based, which will be discussed below. The effective value of the current 20 to be measured is 2.36 times its average value, which results in an energy saving factor of 5.57. A single concentrated current measuring resistor having a value between 0.1 ohm and 0.5 ohm, particularly advantageously having a value of 0.25 ohm, has proven suitable for common power supply devices with a nominal output power of 40 W. However, if the current 20 can be measured very precisely even at very small values, in particular at very small average values, which is essential for the deep dimming of clock-controlled power converters especially in LED operating devices, then the above value becomes low-ohmic. Simply increasing this value would result in unacceptably large measurement losses.

[0105] Due to the lossless susceptance of the admittance, even with reasonable measurement losses, splitting the concentrated current measuring resistor into conductance and susceptance and simultaneously splitting the evaluation into two independent measurements of the average value and the time curve of the current 20 also allows the required precise resolution of the average value measurement. If the admittance formed by the direct parallel circuit of the average value capacitor 10 and the parallel current measuring resistor ( Figures 2a - 2c 11 in Figures 2d - 2f or Figure 1 12 in

[0106] C (10) *R (11或12) has a sufficient time constant, then a voltage proportional to the low-pass filtered value 21 of the current 20, the measured average value, is formed in parallel with the admittance. This time constant must be at least 0.2 times the maximum occurring period duration T of the clock control of the electronic power converter 600 equipped with the current measuring device 100, which has the lowest frequency in its nominal power in the case of a resonant half-bridge. In the example investigated, this is 45 kHz, resulting in a maximum period duration T = 22.2 μs. The size ratio within the admittance is thus calculated as follows:

[0107] However, the time constant of the admittance can also be significantly larger, for example 20 times the above maximum period duration, which represents the maximum value. Therefore, additionally applicable is:

[0108] C (10) *R (11或12) < 444 μs.

[0109] Particularly advantageously, the time constant of the admittance is in the range between 0.6 and 5 times the above maximum period duration, for which applicable is:

[0110] 13.3 μs < C (10) *R (11或12)<111 μs.

[0111] If a clock-controlled power converter equipped with a parallel-connected current measuring resistor has a nominal output power Pnom of 40 W, then values between 1 Ohm and 40 Ohm have proven useful for the parallel-connected current measuring resistor, which provides a share for the admittance. If the power increases, the value decreases accordingly, and vice versa. Thus, its general calculation results in:

[0112] R (11或12) = 40 W * [1 Ohm…25 Ohm] / Pnom

[0113] = 40 V * A * [1 V / A…25 V / A] / Pnom

[0114] = [40 V 2 …1000 V 2 / Pnom.

[0115] A preferred narrower range results in:

[0116] R (11或12) = [80 V 2 …400 V 2 / Pnom.

[0117] For the average value capacitor 10, the general calculation for the range of capacitance results in:

[0118] C (10) = [25…500] * Pnom * ns / V 2 .

[0119] The studied embodiment with Pnom = 40 W includes an average value capacitor 10 with a capacitance of 10 μF and a parallel-connected current measuring resistor of 5 Ohm, i.e., an admittance with a time constant of 50 μs. Generally, this results in for the parallel-connected current measuring resistor:

[0120] R (11或12) = 40 W * 5 Ohm / Pnom = 200 V 2 / Pnom.

[0121] Since the higher the nominal output power Pnom of a power converter equipped with an average value capacitor, the larger this average value capacitor must be, its general calculation is reversed to:

[0122] C (10) = Pnom * 10 μF / 40 W = Pnon * 10 (μAs / V) / (40 V * A).

[0123] Since the Ampere's law cancels out, the general calculation for the exemplary capacitance of the averaging capacitor 10 results in:

[0124] C (10) = 0.25 * Pnom * μs / V 2 .

[0125] In an embodiment, the series current measuring resistor has a conductance value ten times that of the parallel current measuring resistor. A conductance value from one hundred times to twice is possible, and in special cases, the same conductance value is also possible. Preferably, the conductance value of the series current measuring resistor is in the range from twenty times to five times that of the parallel current measuring resistor.

[0126] If the time constant of the admittance is infinite, then Figure 3a the shaded area in the third line graph of Figures 2a to 2f (which corresponds to the loss in the parallel current measuring resistor that contributes to the admittance) is almost as large as the lightly shaded area in the first line graph of the same figure (which corresponds to a similar loss). Since this lightly shaded area is limited by the actual time average 24 of the current 20, if the averaging capacitor 10 from

[0127] In Figure 3b the same parameters (current I with respect to time t and the square of the current I Figure 3a 2 2 ) are shown distributed in the three line graphs in the same scale and in the same way as in

[0128] Here, the area between the curve profile 441 and its time axis is also significantly smaller than the area between the curve profile 400 and its time axis.

[0129] Figure 3c shows the same parameters as Figure 3b before, however now for all power converters except the forward converter, whose converter topology has only one active power transistor, where it operates not resonantly, but rather actually hard-switched, but for this purpose in a so-called critical operation or "critical conduction mode" or "transient conduction mode" or "valley detect mode". The characteristic of this operating mode is that as long as the only power transistor is conducting, the current through the converter inductor increases linearly, and as long as the power transistor is turned off, the current decreases linearly again, and thus, due to the converter topology, the rectifier diode conducts and the energy stored in the inductor is transferred to the converter output. Only when this rectifier diode starts to cut off again (which can be recognized by a sudden voltage change across the converter inductor), the power transistor is turned on again at time point 25'. This is typical for the above-mentioned critical operation, which results in a substantially triangular current 20. Thus, Figure 3c it is applicable, for example, to all boost converters, flyback converters, and SEPICs, and even especially when it operates as a power factor corrector.

[0130] The area difference between the middle line graph and the lower line graph is still always clear here, although not as obvious as in Figure 3a or Figure 3b before.

[0131] In Figure 3d this area difference is even less obvious, Figure 3d shows the power transistor current of a buck converter (CCM Buck) with a non-discontinuous choke current as current 20. Such a clock-controlled power converter is usually used as a post-regulator for compensating voltage fluctuations with minimal losses. "Non-discontinuous (nichtlückend)" or "CCM = continuous conduction mode" already indicates a more continuous current, thereby giving a smaller expression of the area difference between the second and third line graphs: the RMS value and the average value of current 20 approach each other further.

[0132] Conversely, this area difference is in Figure 3a and Figure 3bIt is particularly high because there, the current 20 also includes a negative part. Its square 400 becomes positive and thus increases the square of the effective value. On the contrary, the measured average value 21 of the current 20 is even reduced due to its negative part: thus, the energy-saving factor becomes particularly high.

[0133] In Figure 4a is shown a possible first interconnection formed by a first basic structure 100a of a current measuring device for a clock-controlled power converter according to Figure 2a and an associated regulating circuit 500. The associated adaptation network 150a includes a first series filter 101 which is connected to node 2 and may also include a connection to node 1. Since the first series filter is always coupled directly or indirectly in parallel with the admittance of the current measuring device, as is the case here, the first series filter is a so-called "slow series filter", i.e. mostly a low-pass filter with a time constant which at least corresponds to the time constant of the admittance formed by the mean value capacitor 10 and the first current measuring resistor 11, or may be 50 times or even 500 times larger than the time constant of the admittance. As described above, the current flowing through the connection between node 2 and the first series filter 101 is generally negligible because, according to the classical rules of measuring filters, the input of this first or slow series filter 101 is designed to be as high-impedance as possible. The measurement signal 121 at the output of the first series filter 101 is coupled directly or via a series resistor 141 to the input 211 of the first functional block 205. The measurement signal 121 here corresponds to a further smoothed curve of the measured average value 21 which, as can be seen in the curves 21 of the series in FIG. 3, each has a strong voltage ripple at the clock frequency of the electronic power converter. If there is no direct connection, the series resistor 141 serves for impedance adaptation between the internal electronics of the first functional block 205 and the capacitive output of the first series filter 101 which acts as a slow series filter. In particular, this series resistor can act as an additional but faster low-pass filter in combination with the input capacitance of the first input 211 and prevent very high-frequency interference. Since the resistor 141 is of relatively low impedance and the above-mentioned input capacitance is very small.

[0134] In addition to the admittance formed by the mean capacitor 10 and the first current measuring resistor 11, which serves as the first sensor of the current measuring device for the clock-controlled power converter 100a, a finite conductance exists in the form of a second current measuring resistor 12 between node 2 and node 3 as the second sensor of the same current measuring device. The regulating circuit 500 is designed for this second sensor such that the second functional block 206 of the regulating circuit has a second input 212 and such that the matching network 150a of the regulating circuit includes a second serial filter 102, which can be formed by a direct connection between its input and its output or alternatively by a resistor. Therefore, the second serial filter 102 is also referred to as the "fast serial filter" herein and hereinafter. However, once the second serial filter 102 becomes more complex, it also has a connection to node 1. The second serial filter 102 can also be a low-pass filter, however, which has a time constant between 10 ns and 100 μs, particularly advantageously between 100 ns and 10 μs, differently from the first serial filter that serves as the slow serial filter. The second serial filter 102 thus also remains a fast serial filter. This second or fast serial filter 102 is connected to node 3, which is also referred to as the summing point because at this node, the sum of the instantaneous value 22 of the current 20 and the measured voltage of the measured mean value 21 can be taken as the third measurement voltage 23. The input of the second serial filter 102 is also designed to be as high-impedance as possible according to the classical rules.

[0135] Furthermore, the second serial filter 102 can also be a band-pass filter or a band-stop filter because the third measurement voltage 23 converted by the second serial filter into the second measurement signal 132 has information about time segments that are important for the clock-controlled power converter, from which either the control action, in particular the cut-off time point, for a single power transistor of a simple converter topology can be directly derived or it can be decided whether a more complex power converter (such as a half-bridge) is to be switched off or retracted. The band-stop filter can filter out interference frequencies that may cause direct control action jitter, and the band-pass filter can conduct typical frequencies indicating a fault particularly well to the second input 212 of the second functional block 206, based on which, for example, the switching off of the power converter can be contemplated. All these things must happen quickly, so the second serial filter 102 is always a fast serial filter.

[0136] The output signal of the second-order filter 102 or the second measurement signal 132 (derived from the third measurement voltage 23) is conducted either directly or via the second series resistor 142 to the second input 212 of the second functional block 206. Its task has already been described above for the first-order filter. In the case of a band-pass or band-stop filter, the series resistor 142, in contrast, takes on the task of load adaptation of the second-order filter 102.

[0137] Figure 4b shows Figure 4a semi-integrated or fully integrated or semi-digital or fully digital variants, so that the same current measuring device 100a can be used there. Each order filter can also be implemented by an FPGA or ASIC or by means of software in a microcontroller or microprocessor. The regulating circuit 500 constructed in this way still has two inputs 211 and 212, however, differently from the above, the measurement voltages 21 and 23 are connected to the inputs either directly or only via the series resistors 141 and 142, the possible functions of the series resistors having been described above, and together they form a suitable adaptation network 150b for this purpose. The first input 211 of the first functional block 205 simultaneously forms the input of the "invisible" integrated or digital first or slow-order filter 101b, and its second input 212 simultaneously forms the input of the "invisible" second or fast-order filter 102b.

[0138] In Figure 4c taking into account the fact described above that the third measurement voltage 23 present at the summing point 3 of the current measuring device 100a for a clock-controlled power converter has, in addition to information about the time segments of the current 20 to be measured, the most accurate information about the average value of the same current 20. Since this current is the sum of the measured average value and the measured value for the instantaneous value, where the sum is to be averaged. Therefore, here, differently from Figure 4a the first or slow-order filter 101 is connected to the node 3, having the same task and dimensioning as above, resulting in the only difference between the adaptation network 150c here and the adaptation network 150a above. Compared with the above, the advantage here is a larger measurement voltage amplitude for the same desired regulation result, thereby increasing the signal-to-noise ratio in the functional block, which is illustrated by the measurement signal 123 at the output of the first-order filter 101, which is larger than the measurement signal 121 above by the average instantaneous value of the current 20. This higher signal is coupled to the first input 211 of the first functional block 205 directly or via the series resistor 141 set for a similarly implemented order filter above, the possible function of which series resistor has been described above.

[0139] To evaluate the information on time segmentation, the input of the second or fast sequence filter 102 is connected to node 3 simultaneously here. All other components on this "fast" path are the same as those in Figure 4a in

[0140] Figure 4d shows Figure 4c semi-integrated or fully integrated or semi-digital or fully digital variants. Now, only one connection between node 3 and the only input 210 of the regulating circuit 500 is made either directly or via a series resistor 142, which simultaneously represents the adaptation network 150d. The input 210 thus forms the common input of the first or slow sequence filter 101b and the second or fast sequence filter 102b, and these sequence filters can be integrated or programmed into the regulating circuit 500 as integrated or programmed into the above-mentioned functional blocks 205 and 206, and thus, these sequence filters retain their names there. The bifurcation of the inputs of the two sequence filters can also be implemented by an FPGA or ASIC or by software in a microcontroller or microprocessor.

[0141] The following two figures show the functionally identical interconnections between the current measuring device for a clock-controlled power converter and the evaluation circuit in its functional blocks, once using similar sequence filters and once as semi-integrated or fully integrated or semi-digital or fully digital variants.

[0142] Figure 4e Functionally corresponds to Figure 4c , except that the order of the sensors in the current measuring device is reversed, but this is irrelevant because node 2 is not wired in both figures. In Figure 4e , the second basic structure 100e of the current measuring device according to Figure 2d is connected to the functional blocks 205 and 206 in the same way as in Figure 4c through the already described adaptation network 150c. However, here, the conductance between node 1 and 2 is formed by the first current measuring resistor 11, and the second current measuring resistor 12 between node 2 and 3 contributes to the admittance, the susceptance of which is formed by the averaging capacitor 10, which is also connected between node 2 and 3.

[0143] Figure 4f Functionally corresponds to Figure 4d , except that due to the second basic structure 100e of the current measuring device, the order of the sensors (conductance and admittance) in the current measuring device is reversed, but this is irrelevant due to the unconnected node 2. In Figure 4f , the second basic structure 100e of the current measuring device according to Figure 2d is again in the same way as inFigure 4d is connected in the same way to semi - integrated or fully integrated or semi - digital or fully digital variants of the regulating circuit 500.

[0144] The second basic structure 100e of the current measuring device for a clock - controlled power converter is first connected in Figure 4g by means of two similar sequence filters to the functional blocks 205 and 206. As in Figures 4c to 4f it is also taken into account here that the third measurement voltage at the summing point 3 has the most accurate information about the average value of the current 20. Therefore, for the further smoothing still required, the first or slow sequence filter 101 is now connected to node 3 in order to thereby provide a third measurement signal 123 at its output. As usual, this third measurement signal is passed directly or via a series resistor 141 to the first input 211 of the first functional block 205.

[0145] Here, information about time segments, which is particularly important for power converters (which operate as power factor correctors), is obtained directly from node 2 by means of the second or fast sequence filter 102 (which can also be dimensioned here as described above for Figure 4a ). At this node, there is a second measurement voltage 22 that is proportional to the instantaneous value of the current 20. Since the conductance formed by the first current measuring resistor 11 is located between nodes 1 and 2 here. The fast sequence filter 102 forms a second measurement signal 133 from the second measurement voltage 22, and this second measurement signal is conducted directly or via a series resistor 142 to the second input 212 of the second functional block 206 as described above and, unlike the second measurement signal 132 from Figure 4a does not have the measured average value 21.

[0146] The only difference between the adaptation network 150g here and the adaptation network 150a from Figure 4a above is the crossing of the inputs of the first sequence filter 101 and the second sequence filter 102.

[0147] Finally, Figure 4h shows Figure 4g semi - integrated or fully integrated or semi - digital or fully digital variants of Figure 4bAs described above, the two serial filters 101b and 102b are implemented by an FPGA or an ASIC or by means of software in a microcontroller or a microprocessor. The functional blocks 205 and 206 still have two inputs 211 and 212, where the first input 211 simultaneously forms the input of the "invisible" first or slow serial filter 101b as described above, and the second input 212 forms the input of the "invisible" second serial filter 102b. However, differently from the above, the measurement voltage 23 is connected directly or via a series resistor 141 to the first input 211, and the measurement voltage 22 is connected directly or via a series resistor 142 to the second input 212, which results in a crossing of the measurement lines in the adaptation network 150h.

[0148] List of reference numerals

[0149] 1 First node of a current measuring device for a clock-controlled power converter

[0150] 2 Second node of the same current measuring device

[0151] 3 Third node or summing point of the same current measuring device

[0152] 10 Mean value capacitor

[0153] 10a, 10b Single capacitors which are connected in parallel to form the mean value capacitor.

[0154] 11 First current measuring resistor or its resistance value

[0155] 11a...11d Single resistors which are connected to form the first current measuring resistor

[0156] 12 Second current measuring resistor or its resistance value

[0157] 12a...12d Single resistors which are connected to form the second current measuring resistor

[0158] 20 Current to be measured or its time curve

[0159] 400 Square of the current 20

[0160] 21 Measured mean value or first measurement voltage

[0161] 441 Square of the measured mean value 21

[0162] 22 Second measurement voltage which is proportional to the instantaneous value of the current 20

[0163] 23 Third measurement voltage or sum of the first and second measurement voltages

[0164] 24 Actual mean value of the current 20

[0165] 100* Current measuring device for a clock-controlled power converter

[0166] 101* First or slow sequence filter

[0167] 102* Second or fast sequence filter

[0168] 121, 123 Measurement signals, as output signals of the first or slow sequence filter

[0169] 132, 133 Measurement signals, as output signals of the second or fast sequence filter

[0170] 141 Series coupling resistor at the output of the first sequence filter

[0171] 142 Series coupling resistor at the output of the second sequence filter

[0172] 150* Adaptation network between the current measuring device 100 and the functional blocks 205 and 206

[0173] 203 Input terminal controlled from the outside

[0174] 205 First functional block for implementing current regulation

[0175] 206 Second functional block for implementing overcurrent shutdown

[0176] 210 Only input terminal of the regulating circuit 500

[0177] 211 First input terminal of the first functional block 205 for implementing current regulation

[0178] 212 Second input terminal of the second functional block 206 for implementing overcurrent shutdown

[0179] 300 Clock generator

[0180] 301 Clock signal

[0181] 500 Regulating circuit

[0182] 600 Power component, including a clock-controlled power converter and its current measuring device

[0183] 601 Further first measurement signal from the power component, such as its output voltage

[0184] 602 Further second measurement signal from the power component, such as its input voltage or its critical temperature

[0185] 603 First input terminal for power supply of the power component 600

[0186] The second input terminal 604 is used for supplying power to the power component 600

[0187] The regulated power supply device 700

Claims

1. A current measuring device (100, 100a, 100e) for a clock-controlled power converter (600) for measuring a current (20) in a branch of the power converter, wherein, The current has an effective value, which is at least 1.4 times the average value of the current. It is characterized in that the current measuring device has a series circuit of susceptance and conductance as a sensor, and the current (20) completely flows through both sensors, wherein the current measuring device simultaneously forms at least two different measuring voltages.

2. The current measuring device (100, 100a, 100e) according to claim 1, characterized in that, The average value capacitor (10) forms the susceptance of the susceptance.

3. The current measuring device (100, 100a, 100e) according to any one of the preceding claims 1 and 2, characterized in that, The current measuring device generates a first measuring voltage (21), which is in parallel with the susceptance and is proportional to the low-pass filtered instantaneous value of the current (20), and represents the average value of the current.

4. The current measuring device (100, 100a, 100e) according to any one of the preceding claims 1 and 2, characterized in that, The current measuring device forms a second measuring voltage (22), which is proportional to the instantaneous value of the current (20), wherein the second measuring voltage (22) is in parallel with the conductance, and the current measuring device generates a third measuring voltage (23), which corresponds to the sum of the first measuring voltage and the second measuring voltage.

5. The current measuring device (100, 100a, 100e) according to any one of the preceding claims 1 and 2, characterized in that, The time constant of the susceptance is greater than 0.2 times the maximum occurring cycle duration of the clock control of the power converter (600), and the time constant is at most 20 times the cycle duration.

6. The current measuring device (100, 100a, 100e) according to any one of the preceding claims 1 and 2, characterized in that, The time constant of the susceptance is greater than 0.6 times the maximum occurring cycle duration of the clock control of the power converter (600), and the time constant is at most 5 times the cycle duration.

7. The current measuring device (100, 100a, 100e) according to any one of the preceding claims 1 and 2, characterized in that, The active conductance value of the admittance is related to the nominal output power Pnom of the clock-controlled power converter and corresponds to the resistance value between 40V 2 / Pnom and 1000V 2 / Pnom.

8. The current measuring device (100, 100a, 100e) according to any one of the preceding claims 1 and 2, characterized in that, The active conductance value of the admittance corresponds to the following resistance value, which is related to the nominal output power Pnom of a clock-controlled power converter and results in an ohmic value between 80V 2 / Pnom and 400V 2 / Pnom.

9. The current measuring device (100, 100a, 100e) according to claim 2, characterized in that, The active conductance value of the admittance corresponds to the following resistance value, which is related to the nominal output power Pnom of the clock-controlled power converter and results in an ohmic value between 80V 2 / Pnom and 400V 2 / Pnom, and the capacitance of the averaging capacitor is also related to the nominal output power Pnom and is from Pnom * 25 ns / V 2 to Pnom * 500 ns / V 2 .

10. The current measuring device (100, 100a, 100e) according to any one of the preceding claims 1 and 2, characterized in that, The total conductance value of the conductance is 2 to 100 times higher than the total active conductance value of the susceptance.

11. The current measuring device (100, 100a, 100e) according to claim 10, characterized in that, The total conductance value of the conductance is 5 to 40 times higher than the total active conductance value of the susceptance.

12. The current measuring device (100, 100a, 100e) according to claim 11, characterized in that, The total conductance value of the conductance is 10 to 25 times higher than the total active conductance value of the susceptance.

13. A regulating circuit (500), which is designed to be based on the current measuring device (100, 100a, 100e) according to any one of the preceding claims, has: - A first functional block (205) for implementing the current regulation of the regulated power supply device (700), and the first functional block uses the first measuring voltage (21) as an input signal, and the input signal represents the current at present. - A second functional block (206) for implementing the overcurrent shutdown of the regulated power supply device (700), and the second functional block uses the third measuring voltage (23) as an input signal, and the input signal represents the current at present. It also has an adaptation network (150), and the adaptation network can include a first series filter (101) and / or a second series filter (102), wherein the adaptation network (150) is connected between the measuring voltage and the first functional block and the second functional block.

14. A regulating circuit (500), which is designed to be based on the current measuring device (100, 100a, 100e) according to any one of claims 1 to 12, has: - A first functional block (205) for implementing the current regulation of the regulated power supply device (700), and the first functional block uses the third measuring voltage (23) as an input signal, and the input signal represents the current at present. - A second functional block (206) for implementing overcurrent shutdown of a regulated power supply device (700), the second functional block using a second measured voltage (22) as an input signal, the input signal representing the current at present, further comprising an adaptation network (150), the adaptation network being able to include a first sequence filter (101) and / or a second sequence filter (102), wherein the adaptation network (150) is connected between the measured voltage and the first functional block and the second functional block.

15. The adjustment circuit (500) according to claim 13 or 14, characterized in that, The first sequence filter (101) is a low-pass filter, the low-pass filter having a time constant that is 0.01 times to 100 times the time constant of the admittance.

16. The adjustment circuit (500) according to claim 13 or 14, characterized in that, The second sequence filter (102) consists of a direct connection or a series resistor between a current measuring device (100, 100a, 100e) and an input terminal (212) of the second functional block (206).

17. The adjustment circuit (500) according to claim 13 or 14, characterized in that, The second sequence filter (102) is a low-pass filter, the low-pass filter having a time constant between 10 ns and 100 μs.

18. The adjustment circuit (500) according to claim 17, characterized in that, The low-pass filter has a time constant between 100 ns and 10 μs.

19. The adjustment circuit (500) according to claim 13 or 14, characterized in that, The second sequence filter (102) consists of a band-pass filter or a band-stop filter.

20. The adjustment circuit (500) according to claim 13 or 14, characterized in that, The first functional block (205) and the second functional block (206), as well as the adaptation network (150) having the first sequence filter (101) and the second sequence filter (102), are implemented in at least a partially digital manner in an adjustment circuit (500) as an FPGA or an ASIC with corresponding firmware.

Images