The invention relates to a method for the energy-saving operation of a discharge light source according to the features of the preamble of claim 1. The technical solution presented below is suitable for changing the characteristics of light sources operating on the basis of the discharge principle in an advantageous way.

There have been many efforts in everyday life to reduce the operating costs of low- pressure or high-pressure discharge light sources. Let's highlight among these the solutions disclosed in patent documents HU P1000054 and WO 201 1/092527 where in a circuit arrangement a tapped transformer is interposed to generate specific voltage, current and phase conditions at the terminals of a discharge lamp in a manner that at reduced supply voltage the light source is capable in continuous operation mode to emit less reduced luminous flux than the theoretical flux reduction caused by the voltage reduction.

The shortcoming of this solution has manifested itself in the need for the particularly careful selection of the materials and parts used in the circuit arrangement and the dependence of its efficiency on the physical design and location of the circuit arrangement.

There is still a considerable need for a solution making it possible to ensure the operation of a multitude of discharged light sources, for example the public lighting of an institution or a neighbourhood at less than the original nominal operating costs, efficiently and reliably, if necessary by using cheap devices, suitable for being mass produced, and suitable for complementing the existing systems.

Heated metal emits electrons at a current density that depends on the temperature. Symbolically, one could say that the electrons of the metallic conductivity band are in a potential well, and fill that to the height corresponding to the Fermi energy, the Fermi level. The lip of the well is characterised by zero energy free of external field and corresponding to the potential energy of the free electron. The distance between the lip of the well and the Fermi level is the electron affinity of the metal. Under the effect of heating, a minor proportion of the electrons acquire energy exceeding the Fermi level. In so far as that energy equals the electron affinity, the electron exits the metal. Thermal electron emission depends on absolute temperature (T), the electron affinity (W _KI ) and the material quality of the metal:

In this correlation called Richardson-Duschman formula, AO is the constant typical of the quality of the material. The current density of the thermal electron emission is determined essentially by the W _KI electron affinity and the temperature in the exponential coefficient. Alkali metals are characterised by small electron affinity, hence the surface of a cathode fumed to obtain electrons is coated by such metal.

The degree of the excitation is determined according to the defined energy level (radiation spectrum corresponding to the energy difference between the shells) by the energy needed for the ejection of the electron.

Gas discharge is caused by the potential difference suitable for making the charge carriers move in case of the presence of free charge carriers in the gas, as can be seen in Figure 1. In the known way, distinction is made between a non-self-sustained discharge section 1A, a self-sustained discharge section IB and, within the latter, a transitional range 1C, a glow discharge range ID and an arc discharge range IE.

In a discharge tube 2, electrons 3, while moving from a cathode 4 towards an anode 5, meet gas atoms and as a result photons are generated according to the sketch in Figure 2. With phosphor applied onto the wall of the discharge tube 2, the photons emit light in the visible range. The more photons are generated, the bigger the size of the luminous flux, the illumination level improves. The objective is to generate the highest number of photons.

Traditionally, the number of photons can be increased in as many as three ways:

By increasing the number of electrons 3. The number of electrons 3 provides for current intensity per time unit, that is, the flux can be increased by increasing the current intensity. Raising the current intensity is limited due to the functional properties of a fluorescent tube. The higher the current intensity, the bigger the output and, consequently, with time, also the electric consumption. Current increase can be achieved by increasing the supply voltage.

By raising the supply voltage, the speed of electrons 3 increases, and higher speed generates more collisions and more photons. Higher voltage is concurrent with higher current intensity.

By forming the surface of the electrode 2. The external sealing layer of the electrodes 2 is a result of long research and development activities. According to the experience so far and our current knowledge, in already manufactured light sources it is impossible to change the mentioned properties of the electrodes 2.

Fluorescent tubes can be characterised by efficiency, that is, the quotient of the luminous flux and the output yields the specific indicator expressing the efficiency of energy utilisation in dimensional unit lm/W. This expresses the size of the luminous flux produced at energy of 1 W. An object of providing the method according to the invention was to increase the photometric efficiency of a discharge light source, e.g. fluorescent lamp. In other words, we increase the size of the luminous flux that can be produced of 1 W electric output. Since the number of electrons 3 cannot be increased, the efficiency of the system shall be improved. The invention is based on the recognition that by reducing the electron affinity of electrons 3 the liberated energy increases the speed of the electrons 3, and hence an almost identical quantity of light with fewer electrons 3 can be generated.

With this method, it can be achieved that the electrons acquire more energy than before, thus they detach from the electrode easier and move at higher speed within the discharge tube, despite the fact that the power supplied to the discharge light source is decreased.

In the following, the invention will be described in more detail with reference to the attached drawing, showing an exemplary embodiment of the proposed method. In the drawing, Figure 1 presents the elementary diagram of gas discharge,

Figure 2 outlines the principle of operation of a discharge lamp,

Figure 3 shows the schematic drawing of an exemplary embodiment of a device suitable for the implementation of the method according to the invention,

Figure 4 presents a preferred implementation of the proposed method with the help of a circuit diagram,

Figure 5 shows a further preferred implementation of the proposed method with the help of a circuit diagram,

Figure 6 shows a further preferred implementation of the proposed method with the help of a circuit diagram,

Figure 7 shows a further preferred implementation of the proposed method with the help of a circuit diagram,

Figure 8 shows a further preferred implementation of the proposed method with the help of a circuit diagram, Figure 9 shows a possible solution for eliminating the capacitive character generated by the method,

Figure 10 shows the measuring points used during the examination of the method according to the invention,

Figure 11 shows the course of the voltage and the current intensity in function of time at a chosen measuring point,

Figure 12 shows the measurement values measurable in the various energy-saving settings in the function of time,

Figure 13 shows the current conditions of the presented device in the function of time,

Figure 14 shows the change of the functional properties of fluorescent tubes, Figure 15 shows the change of the functional properties of high pressure light sources,

Figure 16 shows an example of the spectral distribution of the luminous flux of fluorescent tubes and

Figure 17 shows an example of the spectral distribution of the luminous flux of high- pressure lamps. Figure 3 shows an exemplary embodiment of the transformer of a device suitable for the implementation of the method according to the invention. The device comprises a toroidal-core transformer 6 having primary coil 8 and secondary coil 9, called switching coil and actuator coil, respectively, both arranged on a core 7. The primary coil 8 has terminals PI and P2 and branch terminals P3-P6, the secondary coil 9 has terminals SI and S2. The transformer 6 is arranged in the exemplary device implementing the method according to the connection diagram of the basic embodiment presented in Figure 4. The terminal SI of the secondary coil 9 is connected to a phase terminal F of the supply voltage - in the present case mains, 230 V AC -, its terminal S2 is connected to one of the terminals of a discharge light source 10 shown symbolically. The other, neutral or ground, terminal N of the supply voltage is in direct galvanic contact with the other terminal of the light source 10. The terminal PI of the primary coil 8 is connected via ballast 11 commonly with the terminal S 1 of the activator coil 9 to the terminal of phase F, and connected to one of its terminals P2-P6 through a switch unit 12 to the neutral terminal N. In the present example, the tapping switch unit 12 has five positions, in line with the P2-P6 terminals of transformer 6, indicated in the figures by the usual representation of switches in circuit diagrams, switch positions 121, 123, 125 are also marked with reference signs. The circuit arrangement shown enables realising the declared objective, namely using an excitation of the primary coil 8 of the transformer 6 below the rated value. In the simplest case, this can be achieved by the setting of the switch unit 12, that in the switch position 121 provides the smallest, and in the switch position 125 the biggest energy savings. For example, the primary coil 8 of the transformer 6 is rated for 230 V, its secondary coil 9, on the other hand, is rated for a value around 20-30 V. From fifth to eighth of the current flowing through the secondary coil 9 flows through the primary coil 8 in the main branch.

The role of ballast 11 is also to reduce the voltage of the primary coil 8 hence excitation of the toroidal-core 7 of the transformer 6. The ballast 11 is interposed between terminal PI of the primary coil 8 and the phase F, and the ballast 11 may be an inductance or a further transformer. Where, as is visible also in Figure 4, ballast inductance is applied, that may have two terminals, as in Figure 4, but it may also have more terminals branching at various coil positions by the aid of a separate switch 13 as is shown in Figure 5.

If a ballast transformer is applied, then its secondary coil terminals are connected to the terminal PI of the primary coil 8 and to the phase F. The voltage supplied to the primary coil 8 and thereby indirectly the luminous flux of the one or more light sources 10 and hence energy saving is altered through the load connected to the primary coil of the ballast transformer. In the method applying exclusively one transformer 6, the excitation of the transformer 6 is reduced by the number of windings of the primary coil 8 so that an excitation of 100% occurs only connected to the mains, most often to a voltage exceeding the mains voltage. For example, in a possible embodiment, the excitation of transformer 6 will be 100% if 304.8 V is supplied to the central, i.e., third terminal P4 of the primary coil 8. If the mains voltage is connected to this coil part of the primary coil 8 while in operation, the rate of the excitation of the transformer 6 will drop to a value of 75.5%.

The exemplary excitation of 75.5% may also be achieved in another way. The excitation of the primary coil 8 can also be decreased by supplying a voltage differing from the mains voltage (e.g. 230 V in Europe) to its terminals P1-P6. To achieve that, a ballast 11 is connected in series to the primary coil 8. This ballast 11 connected in series to the primary coil 8 acts as a voltage divider. In such case, the primary coil 8 is rated for the mains voltage, e.g. 230 V; nevertheless, it is not supplied with the mains voltage but with less than that. If 75.5% of the mains voltage is connected to the primary coil 8, same excitation can be achieved as in the previously described solution, easiest by a voltage dividing inductance reducing the voltage by 24.5 V.

The two solutions mentioned above can also be used jointly, for example, a primary coil 8 that is rated for a voltage of 268.6 V but is supplied with mains voltage, i.e. 230 V, receives an excitation of 85.6%. In order to attain the previously described value of 75.5%, a voltage dividing inductive ballast 11 is applied. 268.6 V * 0.755 = 202.8 V, that is, if it is supplied with a voltage of 202.8 V, the excitation of the primary coil 8 will have a value of 75.5%. That is, a voltage of 27.2 V will fall on the ballast 11 and a voltage of 202.8 V will fall on the primary coil 8.

The use of the voltage dividing inductance ballast 11 is warranted by the fact that less windings are needed for the transformer 6 and it is easier to manufacture it with less windings. Furthermore, the transformer 6 warms up to a smaller extent for, the longer the wire making up the coil, the higher the number of windings, the more heat is generated. The heat being generated is distributed to the transformer 6 and ballast 11 and, consequently, the thermal dissipation of the device will be better.

Ignition facilitation

During the implementation of the proposed method, a momentary large current surge is generated with the device connected in front of one or more light source(s) 10 upon switching on in feeding direction. Therefore, ignition facilitation is recommended for higher powers of several kW. This task can be accomplished e.g. in the way shown in Figure 6, by interposing a thermal resistor 14 having negative characteristics (NTK/NTC) at an adequate place. At the moment of start-up, the thermal resistor 14 has a resistance of 5-10 Ω, after a few seconds, its resistance decreases to near 0 Ω. This way any negative effect caused by the start-up surge can be prevented. Supply of the light source 10 starts "smoothly", without any excessive surge.

The strength of the current flowing through the thermal resistor 14 is from fifth to eighth of the current strength flowing in the main branch. For a device of an output of 10 kW the effective current flow in the main branch is 43 A. This is reduced by the device at a savings of 40% to a value of 31 A. The peak current value is 44 A. Peak current of 8.8 A flows through the thermal resistor 14, so the type shall be chosen accordingly. At savings of 35% a current of 9.5 A flows through the thermal resistor 14, that is, applying a thermal resistor 14 with 10 A current intensity is adequate. Once the thermal resistor 14 has accomplished its task, it can be eliminated from the circuit, to save it from wear. The elimination may be done e.g. by short-circuiting in a way shown in Figure 7, where short-circuiting is ensured by a mechanical relay 15. After switch-on of the supply voltage, the relay 15 switches in on-position in 8-10 s. The relay 15 can be controlled in the known way by a timer stage, e.g. a closing delay switch 16.

Changing the energy-saving setting during operation

The intended setting is selected by connecting the appropriate terminals, i.e. the respective terminal P2-P6 of the primary coil 8 to the neutral terminal N. Figure 8 shows a possible embodiment. The changeover during operation is feasible by interposing, as for ignition facilitation, e.g. a thermal resistor NTK/NTC 17 in the circuit of switch unit 12, that is, between the respective terminal P2-P6 and neutral terminal N. In this case, no thermal resistor 14 is interposed in the phase F circuit, instead, ignition facilitation takes place in the zero branch. While in operation, thermal resistor 17 has low inner resistance, but the thermal resistor 17 associated with the other saving settings is idle and hence of high internal resistance. In case of a switch-over such a thermal resistor 17 of high initial resistance will be activated. When the switchover is completed, the thermal resistor 17 has a high internal resistance at the moment of the switching preventing formation of any major surges. Due to the current flowing through it, its resistance is soon reduced and the operating state is attained. Thus, a switch-over takes pace smoothly, without surge. A protection circuit for the thermal resistor 17 can be used here as well, of course in a number corresponding to the saving settings. Such a protection circuit comprises a relay 18 short-circuiting the thermal resistor 17 in a controlled way, as well as a closing delay switch 19 controlling said relay 18. In the presented example, there are five thermal resistors 17 and associated protection circuits for the five stages. Of course, as in the case of ignition facilitation, the protection circuit indicated above may be omitted.

Compensation

As a result of the proposed method, a light source 10 and the device feeding it become capacitive. This needs to be compensated for, so that the cos (φ) value should be on the inductive side, close to 1, but at least of a value of 0.9.

As can be seen in Figure 9, a compensating element 20 that may be any inductive element known and used in the art, must be connected on the side of the device implementing the method that is associated with the one or more light sources 10, parallel with the one or more light sources 10. Preferably, commercially available inductances should be used as compensating element 20. Our experience is that the inductive ballasts sold for high-pressure sodium lamps are excellent for this purpose. If cos(cp) remains on the capacitive side, the effect attainable by the method will be less and the device will warm up. Therefore, said correction is absolutely necessary.

In the following, some numerical examples of the effects attainable by the method according to the invention will be shown. The measured values had been measured by clamp meter at measuring points indicated in Figure 10. Measurements are to be made by the clamp meter at the following places:

Measuring point A: main branch, next to terminal SI of the secondary coil 9; Measuring point B: next to terminal PI of the primary coil 8; Measuring point C: next to terminals P2-P6 of the primary coil 8;

Measuring point D: main branch, next to terminal S2 of the secondary coil 9; Measuring point E: next to terminal phase F.

In case of noise-free sine waves without any harmonic content the difference of the current strengths measured at measuring points A and B exceeds the current strength measured in measuring point E. The angular separation of the vectors of the current strengths measured at measuring points A and B is 180°. Where the power supply has harmonic content, said angle of 180° will be offset and the difference of the two current strengths does not result in the input current strength. The bigger the offset, the bigger is the polluting effect. Figure 11 is a time-voltage chart showing the time diagram of the input voltage and the output voltage at a chosen measuring point and, with the help of curves 21 and 22, the measuring values measurable in the various saving settings of the exemplary embodiment shown in Figure 12 are shown in function of time. A curve 23 shows the input voltage, a curve 24 shows the voltage measurable in an active first stage, i.e. using the terminal P6, a curve 25 shows the voltage measurable in the second active stage, that is, using the terminal P5, a curve 26 shows the voltage measureable in the third active stage, i.e. using the terminal P4, a curve 27 shows the voltage measurable in the fourth active stage, i.e. using the terminal P3 and a curve 28 shows the voltage measurable in the active fifth stage, i.e. using the terminal P2. Figure 13 depicts the current conditions of the device being presented in function of time. A curve 29 shows the changes of the input current, a curve 30 shows the changes of the current flowing through the secondary coil 9, a curve 31 shows the changes of the current flowing through the primary coil 8 in function of time.

Figure 14 and Figures 15 show the decrease of the luminous flux of a fluorescent tube, a high-pressure sodium lamp, and a metal halide lamp, respectively, in function of operating time, in case of traditional mains supply and in case of a supply according to the method according to the invention, respectively. A curve 32 depicts a typical decrease of the luminous flux of a fluorescent tube and a curve 33 shows a typical decrease of the luminous flux of a fluorescent tube operated by the method according to the invention. A curve 34 shows a typical decrease of the luminous flux of a metal halide lamp, a curve 35 shows a typical decrease of the luminous flux of a high-pressure sodium lamp and a curve 36 shows a typical decrease of the luminous flux of both of said lamps operated by the method according to the invention: it is easy realisable that the luminous flux remains at an approximately constant level. Figure 16 shows the spectral distribution of the relative luminous flux intensity of a fluorescent tube in function of wavelength, operated by traditional mains supply and by the method according to the invention, respectively. Figure 17 shows the spectral distribution of the relative luminous flux intensity of a high-pressure sodium lamp in function of wavelength, operated by the traditional mains supply and by the method according to the invention, respectively. As can be seen, the proposed method affects the spectral distribution of the relative flux intensity of the supplied light sources 10 to an almost totally imperceptible degree.

The voltage conditions in the specific examples concerning the method according to the invention are as follows: U _N Mains voltage: different by area and country, rating done for 110 V and for 230 V.

U _B Base voltage: the calculations are based on the voltage where savings are smallest and the switch coil has the lowest voltage. The higher the base voltage, the bigger the effect, this is one of the general rules. It is recommended to have at least 1.12 times the mains voltage U _N or more. Thus at U _N = 230 V mains voltage, the base voltage % is 258 V.

U _c Voltage decrease: U _c =47 V at the base voltage U _B .

U _L Voltage jump: having a value between 2.5 AND 5.0 V. U _mjn Minimum voltage U _max Maximum voltage A design example of a device for a mains voltage of 230 V:

Base voltage U _B = 240 V, for feasibility reasons, it cannot be much more than the U _N . At U _B = 240 V the coils can be still arranged in a room available.

Voltage decrease U _C = U _B /5A = 47 V; Voltage jump: U _L = U _B /4S = 5 V

If U _N = 240 V then U _N = U _B → U _L = 5 V

If U _N = 230 V then U _L =U _N /U _B → U _L = 4.8 V

If U _N = 220 V then U _L = U _N /U _B → U _L = 4.6 V

Coil voltages: At an input base voltage of 230 V, the output voltage decreases in the respective energy saving settings are as follows:

At terminal P6: 45 V;

At terminal P5: 40 V;

At terminal P4: 35.5 V;

At terminal P3: 31 V;

At terminal P2: 26 V.

The respective coil voltages are as follows: 240 V 45 V;

47/42 * U _B = 268.6 V for 40 V;

47/37 * U _B = 304.8 V for 35.5 V;

47/32 *U _B = 352.5 V for 31 V;

47/27 *U _B = 417.8 V for 26 V.

A design example of another device for a mains voltage of 230 V:

Base voltage U _B = 258 V. The base voltage could be increased by decreasing the voltage jump to a value of 3 V at a mains voltage U _N = 230 V.

Voltage decrease: U _C = U _B /5.5 = 47 V;

Voltage jump: U _L = U _B /77 = 3.3 V

If U _N = 240 V, then U _N = U _B → U _L = 3. 1 V If U _N = 230 V, then U _L = UIU _B → U _L = 3 V

If U _N = 220 V, then U _L = U/U _B → U _L = 2.9 V

Coil voltages:

At an input base voltage of 230 V, the output voltage decreases in the respective energy saving settings are as follows:

At terminal P6: 42 V;

At terminal P5: 39 V;

At terminal P4: 36 V;

At terminal P3: 33 V;

At terminal P2: 30 V.

The respective coil voltages are as follows:

U _B =258 for V42 V;

U=276 V for 39 V:;

U=299 V for 36 V;

U=326.6 V for 33 V;

U=358.8 V for 30 V.

A design example of a device for a mains voltage of 110 V:

Base voltage U _B = 130 V. The higher the base voltage, the better is the effect. For feasibility reasons, it cannot be much more than U _M - At ½ = 130 V the coils can be still arranged in a room available

Voltage decrease: U _C =U _B /4.5 = 28.9 V

Voltage jump: U _L = %/48 = 2.7 V

If U _N = 130 V, then U = U _B → U _L = 2.7 V

If U _N = 120 V, then U _L = U/U _B → U _L = 2.5 V

If U _N = 115 V, then U _L = U/U _B → U _L = 2.4 V

Coil voltages:

At an input base voltage of 130 V, the output voltage decreases in the specific respective energy saving settings are as follows: At terminal P6: 28.9 V;

At terminal P5: 26.2 V;

At terminal P4: 23.5 V;

At terminal P3: 20.8 V;

At terminal P2: 18.1 V.

The respective coil voltages are as follows:

130 V for 28,9 V;

28.9/26.2*U _B = 143.4 V for 26.2 V;

28.9/23.5*U _B = 159.9 V for 23.5 V;

28.9/20.8*% = 180.6 V for 20.8 V;

28.9/18.1 *% = 207.6 V for 18.1 V.

The method according to the invention has inter alia the advantage of modifying certain operating properties of the fluorescent tubes by the application of the method.

The power consumption ratings of fluorescent tubes generally and usually range from 4 W to 80 W, this value can be reduced by 30-50% where the method is applied.

The luminous flux of fluorescent tubes generally and usually ranges from 200 lm to 7000 lm, quasi-similar values can be attained when applying the method.

The efficacy of fluorescent tubes generally and usually ranges from 50 lm/W to 90 lm/W, this value can be increased by 30-60% where the method is applied. The life span of fluorescent tubes generally and usually ranges from 10 kh to 15 kh, this value can be increased to 30-50 kh where the method is applied.

The colour temperature of 2700-6500 K remains unchanged where the method is applied.

The colour rendering value of 65-85 (phosphor-dependent) remains unchanged where the method is applied.

The line spectrum remains unchanged where the method is applied.

Properties of high-pressure sodium and metal halide light sources are also modified by the application of the method according to the invention.

The power consumption ratings of high-pressure sodium and metal halide light sources can be reduced by 20-40% where the method is applied.

The luminous flux of high-pressure sodium and metal halide light sources is stationary where the method is applied.

The efficacy of high-pressure sodium and metal halide light sources generally and usually ranges from 50 lm/W to 90 lm/W, this value can be increased by 10-30% where the method is applied

Their life span of high-pressure sodium and metal halide light sources generally and usually ranges from 10 kh to 20 kh, this value can be increased to 20-40 kh where the method is applied. A perceptible blueshift effect occurs in the colour temperature of high-pressure sodium and metal halide light sources where the method is applied.

As can be seen, significant energy savings are accompanied by the increase of both the photometric efficiency and the life span of the light sources.

One electron generates several photons, thus, a smaller current intensity is enough to generate an equal amount of light. Less power means less electrical power and less consumption. According to our measurements, energy saving up to 30-60% can be attained.

Electrons flowing from the electrodes damage the electrodes to a smaller degree and hence increase the life-span of the light sources significantly, 2-3 times according to our empirical experience, reducing thereby the maintenance costs to half or one third.

With the help of the method according to the invention, the emitting capacity of the electrodes can be changed - increased - also in existing light sources without dismounting or converting them and the electrodes will preserve this new and positive property as long as the light source is operated by the method according to the invention. Our observations suggest that the electrodes lose the mentioned changed properties after a while when returning to traditional operation.

Thanks to the method, the structure of the electrodes in the light sources changes, albeit some 50-500 hours are needed for the alteration according to our experiments. The alteration is indicated by the capacity change of the light source accompanied by the increase of the luminous flux. Another advantage of the proposed method is that light sources considered used up also regenerate, their efficiency increases and their luminous flux may improve even up to their novel state.

Based on our experiments it has turned out to be advantageous if the load equalled at least 50%, but did not exceed 100%, of the rated output of a device carrying out the method according to the invention. It was found that the optimum load correspond to 60-90% of the rated value.

The advantageous properties of the method have manifested themselves clearly in every case if the external, cylindrical surface of the device was covered by a copper foil and grounded. The device itself may be fixed by aluminium or copper elements, there must be no magnet or ferromagnetic material nearby. Preferably, a distance of at least 10 cm should be left between several installed devices.

List of reference signs:

1A, IB section 15 relay

1C, ID, IE range 16 delay stage

2 discharge tube 17 thermal resistor

3 electron 18 relay

4 cathode 19 delay stage

5 anode 20 compensating element

6 transformer 21-27 curve

7 core 121 position of the switch

8 primary coil 123 position of the switch

9 secondary coil 125 position of the switch

10 light source P1-P6 terminal

11 ballast SI, S2 terminal

12 switch unit F phase

13 switch N neutral terminal

14 thermal resistor A, B, C, D, E measuring point