WO2018041655A1

WO2018041655A1 - Method for operating a flashlight, and flashlight operating according to the method

Info

Publication number: WO2018041655A1
Application number: PCT/EP2017/071034
Authority: WO
Inventors: Walter Vollenweider
Original assignee: Siemens Schweiz Ag
Priority date: 2016-08-29
Filing date: 2017-08-21
Publication date: 2018-03-08

Abstract

The invention relates to a method for operating a flashlight (10) and to a flashlight (10) operating according to the method, said flashlight comprising a light-emitting element (12, 12'), a voltage transformer (20) and, on the output side of the voltage transformer (20), a storage capacitor (22) acting as a stored energy source. For the optimised definition of a switching frequency and an on time of the voltage transformer (20), an input voltage applied to the voltage transformer (20), an output voltage of the voltage transformer (20) and thus the voltage across the storage capacitor (22) and a predefined or predefinable power to be output are taken into account.

Description

description

Method for operating a flashing light and after the

Procedure working flashing light

The invention relates to a method for operating a flashing beacon and a flashlamp operating according to the method, which for implementing the method and for driving an optical signal element (light-emitting element) of the flashing beacon has a flashlamp circuit hereinafter referred to as a driving circuit. A flashing light is used, for example, to warn people in the event of a fire or the like. Flashing lights and their drive circuits are known in many forms. For example, in aircraft at the airfoil tips and / or at the tail unit are known

Flashlights used as collision warning lights. The trigger circuit controls the respective signal element and determines the duration of the light and dark phases. Another application is flashing lights, which are used to identify danger spots, such as construction sites.

The state of the art may be referred to by way of example to US Pat. No. 4,687,081, US Pat. No. 6,822,400 and US Pat. No. 7,663,500.

Flashing lights must deliver a very high light output within a very short time. For this purpose, the drive circuit comprises a voltage converter, which generates a sufficiently high operating voltage for the flashing light during operation from an input-side supply voltage.

Usually, a repetition frequency of the light signals (flashes) emitted by the flashlamp lies in the range of 1 Hz. The electrical energy received by the light-emitting element during a flash may be in the range of a few Ws (watt seconds). When used as signal element / light-emitting If the element uses a flash tube, a duration of a flash is well below 1 ms. The current power can be in the range of a few kW. If, on the other hand, LEDs are used as light-emitting elements, a smaller instantaneous power, for example a power in the range of 10 W to 100 W, is sufficient. To deliver the required light power, the flash duration must be correspondingly longer. The duration of a flash is usually in the range of 10 ms or slightly higher for LEDs.

Flashing lights are usually powered by long leads and a relatively low voltage for safety reasons. It is not possible to transmit the peak power required by the light-emitting element via the supply lines. Rather, a cache is needed. This is part of the drive circuit and is located on an output side of a, for example, as a step-up converter (boost converter) acting part of the drive circuit. This buffer stores the power required during a flash and is then recharged as continuously as possible via the supply line.

As a buffer, a capacitor (storage capacitor), in particular an electrolytic capacitor used. A charging voltage to which the storage capacitor, also referred to as a charging capacitor, is charged is determined by the properties of the respective emitting element. For a flash tube, a capacitance of the storage capacitor in the range of C = 100 yF is common, which is charged to U = 300V. The resulting energy is in these Zahlenwer ^¬ th 4.5 Ws (E = CU ^2/2). Typically, the Speicherkon is ^¬ capacitor discharge until the arc is extinguished in the flash tube. This is the case at about 50 V, and the remaining energy is 0.125 Ws. The energy that is removed from the capacitor during a flash Speicherkon ^¬ is reacted all ^¬ recently only to a small extent in light. The larger part is converted into heat and is thus lost. In the event of a flash lamp equipped with LEDs, the storage capacitor basically has the same function but different values. Typical are I OOO yF and 60 V charging voltage. It is often desired that the light intensity of the flashes ^¬ te can be adjusted within certain limits. Grund ^¬ addition, this can be done in different ways.

In flashlamps that use a flash tube as a light emitting element, it is common for the storage capacitor to be charged to a voltage that determines the energy converted and thus the light energy output by the flashlamp and the brightness. With a flash tube, the voltage range is usually between 200 V and 400 V.

The nominal supply voltage of a flash lamp is übli ^¬ cherweise 24 V, often a slightly higher supply voltage is selected to compensate for losses. If the flashing light is located a short distance from a control center, the supply voltage can be up to approx. 32V. When the flash lamp is at the end, however, a longer To ^¬ line, the voltage may drop to the flash lamp to about 12V. The flashing light must therefore operate in an input voltage range of approx. 12 V to 32 V. From this input voltage, the storage capacitor must be charged. As already mentioned, in case of a flash tube as a light-shaping element, a charging voltage Zvi ^¬ rule about 200 V and 400 V is required. This voltage level is achieved by means of a so-called boost converter, ie a voltage converter which increases the input voltage to the level of the charging voltage.

It should be noted, however, that the storage capacitor may be completely discharged during startup. The voltage converter must therefore also be able to

To deliver voltage which is smaller than the input voltage. However, the efficiency may be smaller in this case, because this operating state is rarely given. For the realization and dimensioning of a voltage converter for use in a flashlight with the highest possible efficiency, some parameters must be known, namely the voltage supplied to the voltage converter, the voltage at the output of the voltage converter (output voltage) and thus the voltage across the storage capacitor, and the amount of energy to be delivered. It is also necessary to know the natural sheep ^¬ th at least individual essential components of the voltage ^¬ converter, wherein the inductance and the N-channel customarily, in the voltage converter used as a switch

MOS-FET are of particular importance. When all the desired information is available, it can be determined with which repetition frequency the transistor is turned on and how long it conducts and thus the magnetic field is built up in the inductance. The discharge phase, ie the time in which the magnetic field is degraded again and the energy is transferred to the storage capacitor, can not be influenced. The invention has for its object to provide a method for optimized operation of a flashing beacon and a flashlamp operating according to the method.

This object is achieved by means of a method with the features of the independent method claim ^¬ ge. The object is also achieved by means of a flashing beacon having the features of the parallel independent device claim. This is provided, that the flash light means for carrying out the method described here and below with further details including, in particular a drive circuit for driving an optical signal element of the flash lamp, in particular a functioning as an optical Sig ^¬ nalelement and light-shaping element flash tube or as an optical signal element and the light emitting element acting group of LEDs, wherein the drive circuit comprises a control unit and a voltage converter for optimally driving the respective optical signal element. In the interest of readability, the following description it is - continued using a so-called flash tube as a light-giving Ele ^¬ ment - also in the specific part. An LED or group of LEDs as a light-shaping element and the like are always Mitzu read ^¬ and to be regarded as understood by the description provided herein with comprehensive.

Advantageous embodiments of the invention are the subject of the dependent claims. The references used in this case point to the further development of the subject of the Hauptanspru ^¬ ches by the features of the respective sub-claim. You are not functions as a waiver of the right of a selb ^¬ constant, objective protection for the Merkmalskombina- of the related dependent claims to understand. Of

Furthermore, with a view to an interpretation of the claims in a closer specification of a feature in a subordinate claim, it is to be assumed that such a restriction does not exist in the respective preceding claims. Finally, it should be noted that the method specified here can also be developed according to the dependent device claims and vice versa.

An embodiment of the invention will be explained in more detail with reference to the drawing. Corresponding objects or elements are provided in all figures with the same reference numerals. demonstrate

1 a flashing beacon with a voltage transformer, a storage capacitor functioning as an energy store, and a flash tube as a light-emitting element,

2 shows a flashing beacon with a voltage converter, a storage capacitor acting as an energy store and with an LED as a light-emitting element, 3 and

4 shows a voltage converter in accordance with Figures 1 and 2 with white ^¬ direct detail, FIG 5 is a voltage profile across the storage capacitor as well and

Tables for the optimized determination of a switching frequency and a duty cycle of the voltage converter

The illustration in FIG. 1 shows, in the form of a block diagram, a basically known flashing light 10 with a flash tube 12 as a light-emitting element. The flashing light 10 is connected via a line 14 from a control center (not shown) and a branch line (supply line) 16 to a supply voltage. The branch line 16 and the part of the line 14 to the branch line 16 form the supply line of the flashing beacon 10 and the supplied supply voltage is applied as a supply voltage to the flashing light 10.

The flashing light 10 comprises a filter 18 which prevents interference from the circuit encompassed by the flashlamp 10 from reaching the line 14. Subsequent to the filter 18, the flashing light 10 comprises a voltage converter 20, in particular a voltage converter 20 in the form of an up-converter. By means of the voltage converter 20, a storage capacitor 22, which is connected on the output side to the voltage converter 20 and acts as an intermediate memory / energy store, is charged during operation of the flashing light 10. The flash tube 12 is operated with the voltage of the SpeI ^¬ cherkondensators 22 and the output of a Lichtsig ^¬ Nals (flash) is triggered by a trigger voltage. This is done by means of an ignition transformer 24 with which a high voltage is applied to the flash tube 12 in order to ignite it. By means of a control unit 26 24 required control signals are generated for driving the voltage converter 20 as well as for controlling the ignition transformer and the Voltage converter 20 and the ignition transformer 24 supplied during operation.

If, instead of a flash tube 12 LEDs 12 'are used as light-emitting elements and correspondingly an LED or a group of LEDs acts as a light-emitting element, a partly different type of circuit of the flashing light 10 is neces ^¬ dig. This is shown in the form of a block diagram in Figure 2 ge ^¬ shows and the following description refers to the essential differences from the circuit in FIG 1. The filtering means of the filter 18, the voltage conversion by means of the voltage converter 20 and the function of the Speicherkonden ^¬ sators 22nd is as energy storage corresponds to what has been described above in connection with FIG 1, so that reference is made to avoid unnecessary repetition, the corresponding descrip ^¬ bung there.

In contrast to a flash tube 12, LEDs 12 'are much more sensitive components that can only be operated within narrow limits. For example, if a group of ten serially connected LEDs 12 'is used, the total flux voltage of the LED group is about 34 V. This is slightly dependent on temperature and current, but is essentially given. The current will be around 300mA, so the power consumed by the LEDs 12 'will be 10.2W.

A permissible current must not be exceeded because otherwise the LEDs 12 'may be damaged or even destroyed. However, the current must be reached in order to achieve the required brightness. As light-emitting elements of a flashing light 10 functioning LEDs 12 'are therefore usually ^¬ operated with an electronic power source 28. It is fed from the storage capacitor 22 with a voltage which is significantly above the forward voltage of the LEDs 12 '. The current source 28 supplies the required current at the voltage required by the LEDs 12 'in the present operating state. Such a power source 28 is usually cherweise with the help of an integrated circuit. To this end, the circuit LM3421 company Texas In ^¬ struments is well suited. If there is between the storage capacitor 22 and the LEDs 12 ', a current source 28, the charging voltage can be fundamentally ^¬ additionally as high as desired. A high voltage is advantageous because a storage capacitor 22 in the form of an electrolytic capacitor with a high nominal voltage at a given volume - and thus also at a certain price - can store more energy. The charging voltage is, however, be limited ^¬ upward by the allowable voltage of the power source 28th In the case of using the circuit LM3421 reali ^¬ overbased current source 28 is limited to about 60 volts. The minimum required voltage is' gege ^¬ ben by the LEDs 12th In the example mentioned, it must be at the input of the current source 28 clearly above the forward voltage of the LEDs 12 'of 34 V. It can thus be assumed that a minimum charging voltage in the range of about 40 volts.

A typical value for the capacity of the storage capacitor 22 is 1 * 000 yF. When the storage capacitor 22 is charged to 60V, the power is 1.8Ws. When the storage capacitor 22 is discharged to 40 V, 0.8 Ws remains, so that 1.0 Ws can be supplied to the power source 28. This is much less than in the example of the flash tube 12 with 4.375 Ws. However, because LEDs 12 'have a much higher efficiency than a flash tube 12, the smaller amount of energy may be acceptable.

The discharge of the storage capacitor 22 can in the case of a flashing light 10 with LEDs 12 'very easily terminated ^¬ who. The LM3421 integrated circuit has a control input for switching the power source on and off.

The storage capacitor 22 is preferably set to the highest possible regardless of the desired brightness Charge voltage, since the efficiency of the power source 28 is the best.

In the illustration in FIG. 3, in one embodiment the voltage converter 20 is shown as a boost converter together with the storage capacitor 22 connected on the output side without the remaining functional units of the flashing beacon 10 (FIG. 1, FIG. 2) as a quadrupole. An up-converter used üb ^¬ SHORT- an inductor 30 as a central element. It works in two phases. In a first phase the inductor is 30 by closing a switch 32 (under control of the control unit 26) with the input voltage ver ^¬ prevented. It flows a current that increases approximately linearly with time, with di / dt = U / L. U is the applied voltage and thus the voltage on the supply line 16 less any losses that may be present. The energy in the inductance 30 is E = LI ^2/2 at the end of the charging process. Übli ^¬ cherweise is an inductor with a ferrite core ver applies ^¬ 30th This core goes into saturation at a given current - and thus a certain resulting magnetic field. The magnetic field strength can not rise above the Saetti ^¬ supply field strength. If the current is further increased, the supplied power is converted into heat and is lost. For the inductance 30, the current at which saturation occurs is thus of great importance. The winding of the inductance 30 is usually dimensioned so that it can be permanently loaded with this current. But the Be ^¬ carrying capacity of the coil is generally of less importance, since the current does not flow continuously.

For optimized operation of a flashing beacon 10, one will endeavor to make the current in inductor 30 so large that saturation almost occurs. In this case, the inductance 30 is utilized in the best possible way, and a voltage converter 20 operated in this way can be realized with a comparatively small inductance 30, in particular a smallest possible inductance 30, and thus a low-cost or lowest-cost inductance 30. After the maximum current in the inductance 30 has been reached and thus the magnetic field has reached its maximum strength, the switch 32 is opened by the control unit 26 on the basis of a corresponding control signal. According to the rela- hung U = L di / dt, the writes ^¬ be the behavior of the inductor 30, the current can not instantaneously return to zero, otherwise the voltage would be infinitely large. Rather, a voltage will adjust due to the circumstances and the current will drop according to this voltage. In an up-converter, the inductor 30 is given the

Possibility to flow through a diode 34 current into a load, in the case of a flashing light 10 in the storage capacitor 22.

A typical value of the inductor 30 is 100 μΗ and ty ^¬ pischer permissible current is 1.0 A. At an input voltage of 24 V, the current rises with di / dt = U / L = 24 V / 100 μΗ = 240 A / ms at. The maximum permissible value of 4.2 ^¬ he ys enough. The energy in the inductance 30 is then

50 yves. With the opening of the switch 32, the energy is shifted from the inductor 30 into the storage capacitor 22. The instantaneous voltage across the storage capacitor 22 (and the inductor 30) amounts to in the considered here Be ^¬ operating state 200 V. The drops 32 driven by the inductor 30 current after opening of the switch with di / dt = U / L = 200 V / 100 μΗ = 2,000 A / ms from. The current is thus after

0.5 ys be zero. Theoretically, a new load could start after 5.2 ys. In practice, however, the inductor 30 still leaves some time to swing out. In ^¬ is passed, a switching frequency of 100 kHz can be selected so that the next charging process after 10 ys starts. The power consumed is 100 kHz × 50 YWS = 5 W. The ex ^¬ given power is due to the losses somewhat smaller. With regard to the estimation of the losses to be expected in the operation of the flashlamp 10, which is sketched below, the illustration in FIG. 4 shows the essential elements of the voltage converter 20 designed as a boost converter as well as the associated each associated with parasitic elements causing losses.

The inductance 30 comprises an ideal, lossless inductance 36, in series with a resistor 38, and a capacitor 40 in parallel with this series circuit. The resistor 38 simulates at least the ohmic resistance of the winding of the inductance 30. But there are more Ver ^¬ loss sources, which can also be summarized in this resistance 38th The capacitance of the capacitor 40 results because in the winding many wires are present, which are close to each other and are separated only by thin dielectrics ^¬ . Thus, there are, so to speak, many small capacitors which are combined in the capacitor 40.

As inductor 30 comes for the flashing light 10 an induct ^¬ tivity 30 of the type RLB9012-101KL the company Bourns Inc. into consideration. This has the following properties according to the data sheet:

Inductance: 100 μΗ ± 10%

Ohmic resistance: 0.28 Ω max

Resonance frequency: 3.70 MHz min

Parasitic capacity: 18.5 pF

DC load capacity: 1.4 A max

The allowable current is 1.4A higher than the typical value of 1.0A given above. As a result, the inductance 30 will be larger and the cost higher, but the efficiency will be better.

The value of the capacitance of the capacitor 40 can be easily calculated from the Reso ^¬ nanzfrequenz f

(C = 1/4 n ² f ² L).

As a switch 32 for operating the voltage converter 20, for example, a hereinafter sometimes only briefly as Transis- gate 42 designated N-channel MOS-FET used. A use of other transistors is also conceivable. The respective transistor 42 has a series resistor 44 and a Paral ^¬ lelkapazität 46. A FET IRF740 type MOS FET acting as a switch 32 has the following characteristics:

Resistance between drain and source when switched on: 0.55 Ω max

Capacitance between drain and source: 170 pF typ

Drain current: 10 A max

Voltage between drain and source: 400 V max

The permissible current is 10 A, making it much RESIZE ^¬ SSSR than NÖ due to the actual maximum current seems to be tig. The transistor 42 is by no means oversized. If a less efficient transistor 42 is used, although the costs are lower, the efficiency is also significantly reduced. The diode 34 at the output of the voltage converter 20 is, for example, a Schottky diode. The losses in the diode 34 are relatively low and are not considered here. However, if an inappropriate diode 34 is used, the losses can become very large.

In the estimation of the losses, it is assumed that the resistor 38 of the inductor 30 and the resistor 44 of the switch 32 are connected in series. Is likewise ^¬ accepted that the capacitance 40 of the inductance 30 and the capacitance 46 of the switch 32 are connected in parallel, although the capacity 40 are referenced to the positive input voltage and the other capacitance 46 to ground.

The loss in the resulting total resistance R as a function of the instantaneous current I ± _n st is:

P = T ² .P.

rinst inst The instantaneous current dependent on the input voltage Ui _n is:

inst = t · ^U in ^{/ L}

The time required for the charging process is: T = I _max / di / dt = L * I _max / U _in

The resulting due to the total resistance Energiever ^¬ loss is:

^E R = ^PL inst ^dt =! Inst ^{2 "} R dt = J ₀ ^T t ² RU _in L * dt

Furthermore: 0 t ² · R · U _in ² / L ² dt = 1/3 ³ · R · U _in ² / L ² = E _R

Most losses will occur in the loading phase. In the discharge phase, the currents are similar. The discharge takes much less time. Thus, the losses in the discharge phase can be neglected.

At the end of the charging phase, a charge remains on the ^parasitä¬ ren capacity 40. The corresponding energy could be partially recovered with a suitable circuit. Because the effort is relatively large, one will do without in small boosters on it. The remaining, and so ^¬ with lost energy:

E _c = ^ U _out ² · C

The losses depend on the operating condition. It is angenom ^¬ men that this is particularly unfavorable by the input voltage Ui ^¬ _n only 12 volts. The charging time T is at 12 V input voltage 8.33 ys, instead of 4.2 ys at 24 V, the total resistance R is 0.28 Ω + 0.55 Ω = 0.83 Ω. We erin ^¬ formers us that the inductance is 100 μΗ. The instantaneous output voltage U _ou t of the boost converter and thus the voltage applied to the storage capacitor 22 is 200 V. The total capacitance is 18.5 pF + 170 pF = 185.5 pF. The current to be reached at the end of the charging cycle is 1, 0 A.

Thus, the losses within a charging cycle can be calculated as follows:

T 8.33 ys

R 0.83 Ω

At 12 v

L 100 μΗ

E _R 2.3 yWs

U ₀ ut 200 V

c 188 pF

E _C 3.8 yWs

It will be recalled that at the end of the charging cycle 50 yWs of energy is present in the inductor 30. The abge ^¬ estimated losses are slightly more than 12% of it, but with an unfavorable operating state. Actual losses will be slightly higher because some causes were not taken into account. For example, further losses will occur due to the remagnetization in the ferrite. The losses in the parasitic resistors 38, 44 and in the parasitic capacitances 40, 46 are in each case similarly large in a suitably dimensioned voltage converter 20. It will be appreciated that the voltage converter 20 is not too far from the optimum in the dimensions shown. When a smaller FET 42 is used as the switch 32, its parasitic capacitance 46 is smaller, so that the energy remaining therein becomes smaller. The parasitic ^resistance 44 of the FET 42, however, becomes larger, so that the here resulting losses are greater. It is therefore likely that the circuit will have slightly less loss if a slightly smaller FET 42 is used compared to the IRF740.

In the approach described here, the flash lamp 10 is assumed that the voltage converter 20 has been dimensioned so that the best possible Eigenschaf- be achieved th approximately in a typi ^¬'s operating state for the optimized operation. On this basis, it is shown here how the efficiency can be improved if the voltage converter 20 is operated outside the typical operating state. The representation in FIG. 5 shows the temporal voltage curve over the storage capacitor 22. It is assumed that the flashing light 10, as a light-emitting element, has a

Flash tube 12 uses the storage capacitor 22 to be charged to 300V and that the arc in the flash tube 12 goes out when the voltage drops below 50V.

The timing can be divided into four phases. In a first phase TL, the storage capacitor 22 is charged (the switch 32 is periodically closed during the first phase (charge phase) TL with the switching frequency f of the voltage converter 20 and the duty cycle ED and is periodically closed again for each further first phase TL). , There is no complete certainty as to when this phase will end. Thus, a waiting phase is added as the second phase TW1 between the end of this first phase TL and the ignition of the flash. In a third phase TB the

Flash tube 12 ignited and the storage capacitor 22 is discharged in a short time. Thereafter, in a further, fourth phase TW2, the system waits until the voltages have stabilized. It would be wrong already during the flash or UNMIT ^¬ telbar after the voltage converter 20 back on (the switch 32 periodically to close), as may occur as a result of so-called Afterglow, which means that as As a result of the charge carrier still present in the flash tube 12 an undesired current can flow. The switch 32 is thus permanently open in the three phases TW1, TB, and TW2.

The total time between two consecutive flashes corresponds to the sum of the four phases TL, TW1, TB, TW2. In the case of the usual flash frequency of 1 Hz, this time is 1.0 sec (one second).

The second phase TW1 and TW2 the fourth phase should be kept mög ^¬ lichst short, since during this time no current from the line 14 is based. The line 14 is thus not exploited in the best possible way. The duration of the third phase TB results from the characteristics of the flash tube 12 and is usually significantly less than 1 ms. The duration of the first phase TL should be made ^¬ Lich as long as mög to for For the line 14 as well as possible use ^¬.

Note that the voltage converter 20 is influenced only by the input of the FET 42 (by means of control signals from the control unit 26) and not by measurement of the current in the inductor 30, as shown and frequently used in US 4,687,081 becomes. The possibility of influencing the optimization is therefore limited to the FET 42 via its gate connection in the best possible way to control. If a voltage converter 20 for use in a

Flashing beacon 10 is built with the highest possible efficiency ^¬ who should know some parameters, namely in particular the input voltage, the voltage at the output and thus over the storage capacitor 22, and the amount of energy to be delivered. It is also necessary to know the characteristics of the construction ^¬ parts, the inductance 30 and the usual ^¬ as a switch 32 used N-channel MOS-FET 42 from be ^¬ special importance. If all the information you want can be set, with which repetition frequency, the transistor 42 is made to conduct and how long it conducts, and so the magnetic field is built up in the inductance 30. The discharge phase, ie the time in which the magnetic field reduced again and the energy in the storage capacitor ^¬ 22 is passed, however, can not be influenced.

In the simplest case, the three parameters, namely input voltage, output voltage and amount of energy delivered, are constant. It is now possible to set and control of the transistor 42 to optimize, and computational and ex ^¬ perimental methods are possible. At a voltage wall ^¬ ler 20 for use in a flash lamp 10 is out this Vo- reduction but not given. For one thing, one of the parameters, namely the amount of energy to be delivered, depending on the desired brightness and ge ^¬ can be adjusted. On the other hand, the second parameter, namely the voltage at the input of the voltage converter 20, results from the circumstances and is only in the best case during the whole

Charging time constant. Finally, the third Pa ^¬ parameters, namely the voltage at the output and thus on the storage capacitor 22 during charging from a minimum to a maximum value changed.

Many voltage transformers 20 operate at a constant frequency of the signal with which the transistor 42 is driven by the control unit 26. Such voltage transformers 20 take into account a changing input voltage by adjusting the duty cycle so that the energy in the inductance 30 just reaches the maximum allowable value and thus no saturation occurs. A voltage converter 20 of this type must be dimensioned so that it can deliver the largest possible power ^¬ required, so that the storage capacitor 22 is completely charged within the allowable time between two flashes. If a smaller intensity of the flash light is set 10 ^¬ be forced energy is smaller, the power output remains but the same, so that the storage capacitor 22 is charged in a shorter time. The result is that the line 14, via which the energy is supplied, is not uniformly be ^¬ overloaded, so that the transmission capacity is not utilized in the best possible way.

It is possible to take into account the desired intensity of the light signal emitted by the flash ^¬ light 10 and thus the output power by either the Einschaltdau- er is reduced, so that in each cycle a smaller

Amount of energy is stored in the inductance 30, or by the frequency of the voltage converter 20 is reduced.

The efficiency is also dependent on the output voltage of the voltage converter 20, but changes continuously during La ^¬ devorgangs. There is no known voltage converter 20, which takes into account the instantaneous voltage at its output. Similarly, no voltage transformer 20 is known, which input voltage the three parameters input voltage, current training and amount of energy to be delivered ^¬ be taken into account at the same time, to set the switching frequency and the duty cycle of the voltage converter 20 in the best possible way.

With the description presented here, a voltage transformer 20 is proposed which comprises the three parameters mentioned

(Input voltage, instantaneous output voltage and amount of energy to be delivered) taken into account, in particular a Spannungswand ^¬ ler 20, which regularly, especially in each

Converter cycle, the switching frequency f and a duty cycle ED of the transistor 42 determines in the best possible way. A converter cycle begins when the transistor 42 is turned on and ends when the same transistor 42 is turned on again at the beginning of the next conversion cycle. The length of the converter cycle is thus the reciprocal of

Converter frequency. Assuming a typical

Converter frequency of 100 kHz is the length of the converter cycle 10 ys. The length of a typical converter cycle was previously calculated to be 4.2 ms, so that in a load cycle 420 Converter cycles take place. If the parameters f and ED are recalculated in each converter cycle, 420 arithmetic operations are thus required in one load cycle. But it can also genü ^¬ gen several times to determine the best possible parameters within a charging cycle so that the parameters during several consecutive conversion cycles remain the same.

Voltage transformers in the usual design, but without the optimization of the parameters described above, must know the input voltage, the voltage across the storage capacitor and the amount of energy to be delivered in order to function in ^{zweckmäßi ¬} ger form. The proposed optimization ^¬ tion of the parameters thus provides no additional amount of hardware. However, the computational effort required during the loading phase increases. However, modern control computers are so powerful that this additional effort is not important.

Basically it is possible to calculate the switching frequency f and the one ^¬ switching duration ED due to the three ER-mentioned parameters as well as the characteristics of the inductor 30 and the transistor 42nd This bill is extremely complex. To make matters worse, that the properties of the components are known only with insufficient accuracy. Even within the scope of a fundamentally possible simulation, which replicates the function of the voltage converter 20, the desired optimization is hardly possible because the accuracy of such a simulation is limited and the complexity is very high.

In an economically feasible voltage converter 20 are simplified methods for determining the optimum

Switching frequency f and duty cycle ED necessary. For this purpose, three basically equivalent methods are proposed.

In a first embodiment, a prototype of the voltage converter 20 is constructed and the efficiency by modifiers ^¬ countries of the switching frequency f and the duty cycle ED opti- mized. The three parameters input voltage, output voltage and amount of energy to be dispensed are changed in a systematic manner and for each combination the best possible switching ^¬ frequency f and duty cycle ED determined. For each of the two parameters input voltage and output voltage for example, sixteen values are recorded, so that there are 256 combi nations ^¬ (16 ^2). For the parameter quantity of energy to be delivered, four values suffice in a flashing light of the usual type, so that finally 16 ² × 4 = 1024 combinations are possible. The process of determining these values can be automated, so that the fact that the Determined ^¬ development is time consuming relativized. In addition, the Determined ^¬ development of these values has to be done only once. The result is stored in a fixed value table (lookup table), for example a fixed value table with 1024 entries of 8 or 16 bits, or the like, in the control unit 26 of the flashlamp 10.

For this purpose, the representation in FIG. 6 shows in symbolic form the content of such a table 48. With U _in (1) to U _in (k), U _out (1) to U ₀ ut (m) and E (1) to E ( n), the assumed values for the input voltage υ _{± η} , the output voltage U _ou t and the respective predetermined amount of energy to be delivered to obtain the desired light output of the flashing beacon 10 are designated. For each recorded value tuples heard a stored in the table 48 in the same row best Schaltfre ^acid sequence f and duty cycle ED. The table 48 is stored, for example, in a memory (not shown) of the control unit 26. During operation of the flash lamp 10 is _η on the basis of the respective actual values for the input voltage υ _±, the output voltage Uout as well as the respective predetermined ge ^¬ desired light output 48 determines the row of the table with the closest values, where it is the deposited for these values of the parameters switching frequency f and switch-on duration ED. The control unit 26 generates as ^¬ raufhin a control signal for driving the switch 32 of the voltage converter 20 with this switching frequency f and a pulse width corresponding to the duration ED. The tax standardized 26 outputs this control signal to the switch 32 of the clamping ^¬ voltage converter 20 and applied in accordance with the clamping ^¬ voltage transformer 20 with the respectively read switching frequency f and the duty cycle ED also read out.

In a second embodiment, the best possible values for the switching frequency f and duty cycle ED are determined when the three parameters reach their limits or are close to the limits. There are a total of eight combinations possible. If the parameters are within this

Range, the switching frequency f and the switch ^¬ duration ED is determined by a means of the control unit 26 performed linear interpolation. For this purpose, the representation in FIG. 7 shows in symbolic form the content of such a table 48. With Ui _Nm i _n and Ui _Nmax , U ₀ uTmin and UouTmax and E _m in and E _max are the respective minimum and maximum values for the input voltage υ ± _η, the output clamping voltage U _ou ^¬ t and each for obtaining the desired light power of the flash lamp 10 denotes predetermined energy to be delivered ^¬ quantitative. For each value tuples are in the table 48 in the same row stored values for the Schaltfre ^acid sequence f and the duty cycle ED. These values can be called, for example f min / min / min _/ _m ED in / min / min, f max / min / min, etc.

Also, such a table 48 is stored for example in a SpeI ^¬ cher the control unit 26th

During operation of the flash lamp 10 from the output voltage Uout as well as ^¬ each predetermined to be dispensed desired amount of energy to be f and the duty cycle ED determines the values of the switching frequency with linear interpolation using the respective actual values for the input voltage _υ ± η.

The interpolation can be carried out in the best way possible for the arithmetic unit of the control system used. computer of the control unit 26 is suitable. For example, the interpolation can be carried out in the following manner.

Step 1

The output f is set equal to the optimum value for the minimum values of input voltage, output voltage and amount of energy to be delivered, so that:

f ^- fmin / min / min

2nd step

The dependence on the input voltage is taken into account by first determining the slope:

(fmax / min / min fmin / min / min) / (u INmax U INmin I

The optimum switching frequency is determined by multiplying the input voltage Um by the slope and added to the switching frequency fmin / min / min to give: f ^- fmin / min / min

+ U _in (fmax / min / min fmin / min / min) / (u INmax U INmin I

It should be noted that the slope can also be negative.

3rd step

The dependence on the output voltage is taken into account by calculating the slope again:

(fmin / max / min ^- fmin / min / min) / (Uouimax ^- Uouimin)

The output voltage is multiplied by the slope and added to the previously calculated value of the switching frequency: f ^- fmin / min / min

+ U _in (fmax / min / min fmin / min / min) / (u INmax U INmin I Uout (fmin / max / min fmin / min / min) / (UOUT ': max

Step 4 The desired quantity to be delivered energy E is taken into account in gleicharti ^¬ ger manner: f = f min / min / min +

Other equivalent computational methods are possible, especially if there is a desire to divert to divisions.

5th step

The optimum duty cycle ED is determined in a similar manner as the optimum switching frequency f.

If the values of the optimum switching frequency f and duty cycle ED depend strongly on the input variables in a highly non-linear manner, the linear interpolation can lead to large errors. A higher-order interpolation leads to better results. However, it is no longer sufficient to specify only the optimal parameters at the extreme values or in the vicinity of the extreme values of the input variables, but also interpolation points within the value range are necessary. Thus, a table similar to that shown in FIG. 6 but containing fewer values is required.

The methods to be used are much more complicated than in the case of linear interpolation. It is not necessary to describe them in detail here, since they are generally known to the person skilled in the art.

With the determined optimum values for the switching frequency f and the Duty cycle, the control unit 26 as ^¬ f raufhin a control signal for driving the switch 32 of the voltage converter 20 at this switching frequency and a pulse width corresponding to the duration ED. The control inputs standardized 26 outputs this control signal to the switch 32 of the clamping ^¬ voltage converter 20 and applied in accordance with the clamping ^¬ voltage transformer 20 with the respectively read switching frequency f and the duty cycle ED also read out. In a third embodiment, a complete or simplified equation system is used to determine the optimum switching frequency f and duty cycle ED as a function of the three parameters. This system of equations is sufficiently frequently achieved by means of the control unit 26, and in particular in each cycle of the voltage converter 20 or ^¬ Wenig least newly dissolved in a few cycles. The Gleichungssys ^¬ system is stored together with a functionality for releasing the sliding ^¬ surveillance system, for example in a memory of the control unit 26th During operation of the flash lamp 10 is turned on hand of the respective actual values for the input voltage _U, the output voltage U _out as well as each specified ^differently bene desired light output solve the equation system, whereupon the control unit 26 a control signal for dently ^¬ tion of the switch 32 of the voltage converter 20 generated with the respective resulting switching frequency f and a pulse width corresponding to the resulting duty cycle ED. The control unit 26 outputs this control signal to the switch 32 of the voltage converter 20 and acts in accordance with the voltage converter 20 with the respectively calculated switching frequency f and the likewise calculated duty cycle ED.

The three specified methods set the switching frequency f of the voltage converter 20 and the duty cycle ED of Transis ^¬ sector 42 in the best possible manner so that the storage capacitor 22 capacitor just at the end of the allowable time has the required voltage. The voltage converter 20 is acted upon by the control unit 26 with the respective resulting values by the control unit 26 for controlling the Switch 32 of the voltage converter 20 outputs a control signal with the respective switching frequency f and a duty cycle of the switch 32 defining pulse width. It can not be ignored that the three proposed methods have a necessarily limited accuracy. It is therefore possible that the storage capacitor 22 is already charged before the end of the allowable time or that it is not fully charged within this time.

In a specific implementation of the approach proposed here, therefore, an additional device can be provided in each of the three described embodiments, which determines whether the storage capacitor 22 is completely charged within the time provided, ie within the charging phase TL. If this is not the case and the charging process is terminated either too early or too late, a correction value is calculated. This is used in a control loop in order to influence, for example, the value of the desired energy, and subsequently the calculation of the switching ^¬ frequency f and the duty cycle ED so that the storage capacitor 22 is precisely as possible fully charged at the end of the charging phase TL. The calculation of the corrective ^¬ turwerts done, for example by means of the control unit 26th

Although the invention has been further illustrated and described in detail by the exemplary embodiment, the invention is not limited by the disclosed example (s), and other variations can be derived therefrom by those skilled in the art without departing from the scope of the invention.

Claims

2 b claims

1. A method for operating a flashing light (10), which comprises a light-emitting element (12, 12 '), a voltage converter (20) and the output side of the voltage converter (20) acting as a energy storage capacitor ^¬ memory (22), wherein for optimized determination a switching frequency and a duty cycle of the voltage converter (20) an input voltage at the voltage converter ^¬ (20), one at the output of the voltage converter (20) and thus the voltage over the Spei ^¬ cherkondensator (22) applied voltage and finally a predetermined or predetermined energy to be dispensed be taken into ^account .

2. The method of claim 1, wherein a prototype or a model of the clamping ^¬ voltage converter (20) is used for optimized determination of a switching frequency and a duty cycle of the voltage converter (20),

- In which the three parameters input voltage, output voltage and amount of energy to be emitted are changed in a systematic manner ^¬ changed and for each combination the best possible

Switching frequency and duty cycle is determined

- Wherein for each of the three parameters a predetermined or predetermined number of values are recorded and stored in an automatically evaluable table (48) and

- Wherein operation of the flashing light (10) based on the jewei ^{¬ term} actual input voltage, output voltage and the predetermined or specifiable amount of energy to be delivered based on the table (48) each selects the best possible switching frequency and duty cycle and the voltage converter (20) is acted upon.

3. The method of claim 1, wherein for optimal determination of a switching frequency and a duty cycle of the voltage converter (20) the best possible values for the Schaltfre ^¬ frequency and the duty cycle are determined when the three parameters input voltage, output voltage and amount of energy delivered their limits or near the Limits lie, and the resulting values are stored in an automatically evaluable table (48), and possibly also support values are stored within this range from ^¬ , wherein during operation of the flashing light (10) based on a respective actual input voltage, output voltage and the predetermined or predetermined Performance and based on the table (48) by interpolation the best possible switching frequency and duty cycle ^¬ selected and the voltage converter (20) is thus applied.

4. The method of claim 1, wherein the optimized determination of a switching frequency and a duty cycle of the voltage converter (20) a complete or simplified Glei ^¬ chungssystem for determining the optimum switching frequency and duty cycle in dependence on the three parameters is used and wherein the system of equations in the operation of Flashing light (10) on the basis of the respective actual input voltage, output voltage and the predetermined or predetermined power to obtain the best possible switching frequency and duty cycle is released and the voltage converter (20) is acted upon by the resulting values.

5. The method of claim 2, 3 or 4, wherein the best possible switching frequency and duty cycle regularly, in particular re ^¬ in each converter cycle or in at least one charge cycle several times, redetermined

6. The method according to any one of the preceding claims,

wherein it is determined whether the storage capacitor (22) is fully charged within the duty cycle,

- where a correction value is calculated when the charging process ^¬ gang is terminated either too early or too late,

- Wherein the calculation of the switching frequency and the switch-on duration is influenced by means of the correction value or the determined switching frequency and the determined switch-on ^¬ duration are influenced by the correction value so that the storage capacitor (22) as close as possible to the end the duty cycle is fully charged.

7. The method of claim 6, wherein the influencing takes place by means of a control loop.

8. flashing light (10) with means (20, 26) for carrying out the method according to one of claims 1 to 7.

9. Flashing light (10) according to claim 8 for carrying out the method according to one of claims 2 to 7, wherein the control ^¬ unit (26) as a means for determining the ^{bestmögli ¬} chen switching frequency and duty cycle and as a means for acting on the voltage converter (20) with this switching ^¬ frequency and this duty cycle acts.

10. flashing light (10) according to claim 9 for carrying out the method according to one of claims 6 or 7, wherein the control ^¬ unit (26) as a means for influencing the calculation of the switching frequency and the duty cycle based on the correction value or as a means for influencing the determined

Switching frequency and the determined duty cycle on the basis of the correction value acts.