Patent Application
20050104536
The present invention relates to a device and a method for operating a discharge lamp, such as a fluorescent lamp, halogen lamp etc.
Devices for operating a discharge lamps or ballasts are widely used for providing a controlled power supply to the discharge lamp. Typically, power control circuitry controls the lamp driver circuit which comprises a switched-mode power supply (SMPS) connected between the mains and the discharge lamp. The power control circuitry may be employed to optimize the preheating and ignition of the discharge lamp, to maintain a constant power to the discharge lamp for the purpose of maintaining a selected light intensity or may be used for the purpose of controlled dimming of the light intensity of the discharge lamp.
Recently digital devices for operating a discharge lamp or digital ballasts are developed wherein the power control circuitry employ digital techniques for controlling the power supplied by the switched mode power supply to the discharge lamp. Digital ballasts provide a relatively low cost control of the power, voltage and/or current supplied by the power supply. Digital ballasts are versatile as compared to the analog ballasts and allow for easier implementation of complicated control and tiing processes.
For the purpose of output power control a specific type of ballast may determine the values of one or more lamp parameters, such as the lamp voltage, the lamp current, and/or the lamp power, and use the determined values in the control process of the power supply. Consequently, the parameters values are measured and one or more signals representative of the measured parameter values are fed back into the power control circuitry. The power control circuitry uses the parameter signals to control the output voltage, output current and/or output power actually provided to the lamp by the power supply. However, the accuracy of this control process depends inter alia on the accuracy of the determined parameter signals and the sensitivity for errors in these signals.
To improve the control process the parameter signals may be filtered by using one or more analogue filters, e.g. filters including passive elements such as resistors and capacitors.
A drawback hereof is that if analogue filters are applied in a ballast, the characteristics of the filters are dependent on the applied hardware, i.e. on the specific passive elements applied. When in various situations filters with different filter characteristics are needed, the hardware used in a first situation must be replaced by different hardware in another situation.
A further drawback is that the filter characteristics remain constant after the filter is placed in the ballast. This implies that the filter characteristics of the filter cannot in general be changed once the ballast is fabricated. For example, at the end of the life of a particular lamp used, the control of the power supply may require filter characteristics of the control signal filter which differ considerably from the optimal filter characteristics in case a new lamp is used.
A still further drawback is that due to the inflexibility of the prior analogue filters, each of the parameter signals is to be filtered by a separate filter, which requires a considerable number of electronic components and renders the ballast circuitry complex.
A still further drawback is that the analogue filters are unable to adapt the filter characteristics during operation of the ballast. This may for example be needed in case an optimal power supply control during changing of the power supplied to the lamp, such as during dimming of the lamp, requires changing the filter characteristics of the filter(s).
It is an object of the present invention to improve the existing devices for operating a discharge lamp and to provide a device wherein at least one of the above-mentioned drawbacks is obviated.
According to a first aspect of the invention a device for operating a discharge lamp is provided, the device comprising:
According to a preferred embodiment the digital filter is software-controllable. Thus, the operation of the filter, for example the characteristics of the filter, can be easily changed by simply loading an adapted version of the software controlling the filter.
In a further preferred embodiment the filter is adapted so as to control the characteristics of the digital filter during operation of the discharge lamp. The filter characteristics for example may be changed depending on certain predefined values of the measured control signal(s) or may be changed as function of the life of the lamp in use.
In a further preferred embodiment the converting means comprise a first analogue-to-digital (A/D-) converter for sampling a first lamp control signal and a second analogue-to-digital (A/D-) converter for sampling a second lamp control signal. When measuring three or more signals, the converting means may comprise three or more analogue-to-digital converters, one analogue-to-digital converter for each measured control signal. The resulting digital control signals may each be submitted to a digital filter. Preferably, however, each of the resulting digital control signals is filtered in one and the same digital filter, which further reduces the number of electronic components needed to implement the operating device.
In a further preferred embodiment the converting means comprise one analogue-to-digital (A/D-) converter for successively sampling each of the lamp control signals. In this embodiment the various measured analogue control signals are successively sampled by one and the same A/D-converter and consequently the circuit design may be simplified even further.
In a preferred embodiment of the present invention the digital filter is a first order filter, wherein the first order filter preferably processes the digital lamp control signal according to
wherein ON is the filtered digital lamp control signal for time point N, ON-1 is the filtered lamp control signal for time point N-1, IN is the digital lamp control signal on time point N and X is a software-controllable filter parameter and wherein X preferably is a preset integer. A first order filter is relatively simple and the amount of program source code needed to implement a first order filter is limited.
If the need arises for a stronger filter then the digital filter may comprise two or more first order filters in series to create a second order filter and so on. However, in further preferred embodiments the second and higher order filters may be programmed directly.
In another preferred embodiment the digital filter comprises a buffer array for storing of a plurality of input samples of the digital lamp control signal and means for processing at least a part of said plurality of input samples in the buffer array to provide an output sample of the digital control signal. Although application of a buffer array may require a relatively large memory capacity, this embodiment will provide a fast and versatile digital filtering of the lamp control signal.
Preferably the buffer array has a first-in first-out (FIFO) structure, which means that input data samples are stored into an array of a number (N) of entries and that the oldest input data samples are shifted out at the moment a new sample has to be placed into the buffer array. All entries or at least a plurality of entries are used to filter the input data.
In a further preferred embodiment each sample of the plurality of input samples of the digital lamp control signal a different weight factor is applied, whereafter the weighted input samples are summed to provide the output sample of the digital control signal, preferably providing a moving average filter having a sinc-shaped frequency response.
In a preferred embodiment the filter means and control means are implemented in one microcontroller. The microcontroller comprises at least a central processing unit, a memory in which the control software may be loaded, input- and output terminals and interconnecting circuitry. The microcontroller incorporates both the function of control circuitry for the power supply and the function of filter for the control signals used by the control circuitry. Both functions may be implemented by the same software-program running on the microcontroller.
According to another aspect of the invention a method for operating a discharge lamp is provided, comprising the steps of:
Further advantages, features and details are given in the following description of two preferred embodiments of the invention. In the description reference is made to the annexed figures.
The lamp power supply according to a preferred embodiment of the invention is a dutycycle controlled switched mode power supply (SMPS) of the constant frequency pulse width modulation (PWM) type, which uses the same frequency for ignition, normal operation and dimmed operation of the lamp. In the embodiment shown in
The switched mode power supply (SMPS) operates in the symmetrical mode. The dutycyles of the two switching elements are equal, their on-times being separated from each other by ½ of the switching period. In the ignition phase the L-C combination Llamp, Clamp is unloaded which generates a high voltage across the lamp. This causes ignition of the lamp. In the bum phase the L-C combination Llamp and Clamp is loaded by the lamp. The power delivered to the lamp is determined by the dutycycle. Hence, the lamp power supply is controlled by one parameter, the dutycycle of the switching elements.
In the block diagram of
For driving the lamp a half-bridge drive circuit is shown, wherein the switching elements are formed by two power transistors (power FET's) Q1 and Q2. The gates of the switching elements Q1 and Q2 are driven by driver signals GHB1 and GHB2 originating from a control circuit to be described hereafter.
Further are shown an LC-combination Llamp, Clamp for driving the lamp and control circuitry for providing the control signals GHB1 and GHB2 to power transistor Q1 and Q2 respectively. As the control circuitry operates on a relatively low voltage (typical 5 V supply voltage), the input signals must be in the range from 0 to 5 V and consequently the output signals that the control circuitry can deliver are also in this range. Consequently, the control circuitry is provided with an interface circuit (IFC) for converting voltages and currents into usable indication signals and for converting control signals from the control circuitry into usable driver signals for the switching elements Q1 and Q2. The control circuitry is provided with a microcontroller (MC) including read-only memory (ROM), programmable or non-programmable, random access memory (RAM) and/or a processor, A/D-converters, D/A converters etc.. In the memory of the microcontroller control software is stored. Instead of a microcontroller a special purpose digital signal processor (DSP) may be used, which includes a CPU especially designed for digital signal processing. In a DSP extra fast instruction sequences are provided to improve the signal processing performances of the device.
Although not shown in
The control circuitry (1) outputs, under software control, a square wave, which is averaged in the interface circuit with an RC-filter to rule out the ripple component. The resulting DC-voltage is used by the control circuitry (1) to generate the driver signals GHB1 and GHB2 for the switching elements Q1 and Q2 respectively. The driver signals GHB1 and GHB2 may in another embodiment of the invention be generated directly by the microcontroller. A level shifter (not shown) will be used to bring the driver signal GHB1 at the appropriate level. Consequently, the dutycycle, with which the power supply to the lamp is to be controlled, is determined by software stored in the memory of the microcontroller.
The functions of stabilization of the power or current in the lamp, the optimization of the ignition, preheating and electrode heating, the adaptation to different lamp types, can be achieved by adapting the software running on the microcontroller. These functions are implemented by a digital control loop for which the microcontroller performs measurements of a plurality of physical quantities or parameters such as the current in the lamp, the voltage across the lamp, the supply current and supply voltage.
One of the parameters may be the current Ilamp running in the lamp. Ilamp can be determined in various ways. In the embodiment of
Another parameter may be the actual voltage Ulamp across the lamp. Ulamp can be determined in various ways. In the embodiment of
The above mentioned parameters may be determined using relatively fast A/D-converters which are able to perform a high frequency sampling of the relevant parameter signals.
A further parameter may be the supply current Isupply, which is represented by the averaged voltage across the shunt resistor of divider DI. The resulting analogue signal Isupply,meas is representative of the supply current. Also the supply voltage Usupply may be represented by the averaged voltage Usupply,meas from divider DU.
The analogue control signals Ilamp,meas, Ulamp,meas, Usupply,meas and Isupply,meas are fed to the interface controller (IFC) that converts the signals into usable indication signals for the microcontroller. Thereto each of the analogue control signals is converted into a corresponding digital control signal by one or more A/D-converters provided in control circuitry (1). The control circuitry (1) may convert each of the analogue control signals into corresponding digital control signals using a corresponding number of AID-converters, that is one A/D-converter for each the control signal. However, the microcontroller may also be programmed to use less A/D-converters, or even only one A/D-converter in combination with a multiplexer for converting the analogue control signals into corresponding digital control signals.
Once the analogue control signals are converted into a digital form, they are processed by the microcontroller (MC). Each of the digital control signals is filtered by using a digital filter, in this embodiment a software filter.
In general a first order software filter can be described as:
ON=IN-1*kIN-2*k2+ . . . +IN-M*kM
wherein O stands for the output result of the filter, k an arbitrary number between 0 and 1, and IN for the Nth input signal. Implemented in software this yields for a specific type of filter:
ON=(1/X)*IN+(X−1)/X*ON-1
wherein X is an integer. If X is large, then the cut-off frequency will be small. When X is small, the cut-off frequency of the filter will be high.
The step response of a hardware-implemented analogue first order filter is a continuous function:
O=(1−e(−t/RC))
wherein t is the time and RC is constant. In a software-implemented filter the response time depends on X and on the repetition sample speed of the input signal. Suppose the digital filter is implemented as follows:
ON=0,25IN+0,75ON-1
Then the “new” input sample contribution to the result ON is a quarter and the contribution of the “old” output sample to the “new” output signal is ¾. If X is increased from 4 to 8, the contribution of the “new” input sample will be reduced to ⅛. If a “stronger” filter is needed, then two first order filters are placed in series in order to create a second order filter with corresponding second order filter characteristics and so on.
In
In
The present invention is not limited to the above-described preferred embodiments thereof; the rights sought are defined by the following claims, within the scope of which many modifications can be envisaged.
| Number | Date | Country | Kind |
|---|---|---|---|
| 01204611.6 | Nov 2001 | EP | regional |
| Filing Document | Filing Date | Country | Kind |
|---|---|---|---|
| PCT/IB02/04676 | 11/6/2002 | WO |
