DRIVING ELECTROLUMINESCENT DISPLAYS
This invention is concerned with electroluminescent displays, and relates in particular to the manner in which these displays are powered and driven.
Certain materials are electroluminescent that is, they emit light, and so glow, when an electric field is generated across them. The first known electroluminescent materials were inorganic particulate substances such as zinc sulphide, while more recently found electroluminescent materials include a number of small molecule organic emitters known as organic LEDs (OLEDs) and some plastics synthetic organic polymeric substances known as light emitting polymers (LEPs). Inorganic particulates, in a doped and encapsulated form, are still in use, particularly when mixed into a binder and applied to a substrate surface as a relatively thick layer; LEPs can be used both as particulate materials in a binder matrix or, with some advantages, on their own as a relatively thin continuous film.
This electroluminescent effect has been used in the construction of displays. In some types of these a large area of an electroluminescent material generally referred to in this context as a phosphor is provided to form a backlight which can be seen through a mask that defines whatever characters the display is to show. In other types there are instead individual small areas of EL material. These displays have many applications; examples are a simple digital time and date display (to be used in a watch or clock), a mobile phone display, the control panel of a household device (such as a dishwasher or washing machine), and a hand holdable remote controller (for a television, video or DVD player, a digibox, or a stereo or music centre).
The phosphors used in electroluminescent displays require quite a high voltage of the order of 200 volts to drive them, but take only a minuscule current in the microamps region. For hand held devices, therefore, it is
perfectly feasible for them to be driven by conventional low voltage 2 to 9 volt power supplies, in the form of a suitable number of standard 1.5 volt AA batteries plus some appropriate solid state electronics to provide the required 200 volt output voltage. And rather than provide such a voltage in a DC form, it is conventional to "pulse" the display phosphor using an AC voltage with a conventional sine wave form; the display's electroluminescence persists for some time even when the driving voltage drops (during the AC cycle) below an effective level, and this, coupled with the normal human persistence of vision, means the display looks "on" all the time the driving voltage is applied.
Using an AC driving voltage very conveniently allows the brightness of the display to be varied either by varying the peak value of the voltage or by varying the AC frequency; lower peak values, or lower frequencies, cause the display to appear dimmer, while higher ones make it brighter. This results in acceptable control of the display when, as in older displays, the electroluminescent phosphor is turned on or off as a whole, providing its illumination through a mask defining the symbols to be seen by the User, but is less satisfactory when, as in more modern displays, each symbol is generated by an individual patch of phosphor that is activated quite independently of all the other patches. The invention is specifically directed toward the improved performance and control of such multiple segment electroluminescent displays, as is now explained in more detail.
Multiple segment displays are most conveniently driven using switched mode power supplies (typically as described in our British Patent Publication No: 2,372,647). There are many references in the Art to resonant circuits able to achieve voltages high enough to power an electroluminescent lamp. However, there are a number of challenges to overcome when trying to combine the efficiency of a resonant drive topology with the control required for a multiple segment display.
The displays of the Art typically use auto resonant circuits. Circuits of this type oscillate automatically on application of power at the frequency determined by their resonant components. These are usually an inductive element and the capacitance of the electroluminescent load. These circuits are variously described as series or parallel resonant dependant on whether the capacitor and inductor of the resonant circuit are connected in series or parallel. In either case, current flows through the inductor/capacitor combination, and energy is alternately stored as charge on. the capacitor and magnetic field in the inductor. The inductive element itself may be a two terminal inductor, or it may be associated with a transformer winding. In each case, the auto resonant circuits provide positive feedback directly in hardware (by means of transistors and/or amplifiers, sometimes with the use of a phase locked loop) to cause oscillations.
These auto resonant circuits have significant disadvantages for multiple segment applications. It is advantageous to be able to control the brightness of the display in the presence of dynamically changing loads arising from such factors as the segments' changing states (turn on and off either to vary the displayed image or as part of a grey scale control) and the display's ageing. The invention proposes that rather than allow free and uncontrolled resonance, there be used, instead, means typically, a microcontroller to detect the relevant state of the resonant circuit that is, of each electroluminescent segment or group of segments then selected and to keep the circuit under control as the load and display requirements change.
In a first aspect, therefore, the invention provides a drive and control system for a multi segment electroluminescent display, which system incorporates means for driving the segments of the display by an AC voltage, wherein there are programmable sequential state machine means to detect the relevant state of each appropriate resonant circuit, and to keep the circuit under control as the load and display requirements change.
The invention's drive and control system incorporates means for driving the segments of the display by an AC voltage. These means may take any appropriate form, but in general are most conveniently resonant circuits of the type described hereinbefore, albeit with the modification proposed by the invention. However, in outline a specific embodiment is described in more detail hereinafter with reference to the accompanying Drawings the drive and control system of the invention, in its more preferred aspects at least, may generally be described as follows.
A microcontroller is operatively connected between the primary coil of a driving transformer, to which are operatively connected each of a set of illuminatable display segments, and a corresponding set of switches that control whether the segments are on or off. There may be as many as 20 segments, and more; the only limiting factor is the practicality of controlling their switching, which is here limited by the number of the microcontroller's outputs.
According to the invention, the AC voltage display segment driving arrangement includes programmable sequential state machine means to detect the relevant state of the appropriate resonant circuit, and to keep the circuit under control as the load and display requirements change. This means may take a number of different forms, but most conveniently is a microcontroller having inputs from and outputs to each individual resonant circuit, and suitably programmed to effect via the outputs the required control of each circuit based upon these inputs. A typical microcontroller appropriate for this purpose is the PIC16F57 from Microchip Inc, and an example of its programming for this purpose, given as an outline flow diagram, is shown hereinafter.
A microcontroller is an appropriate device to use for this purpose as it allows control of the circuit with a low cost, flexible part/ However, any sequential state machine could be used for this purpose. A very high
volume design might well use a sequential state machine built as a custom (or semi custom) silicon chip.
As will be apparent from the more detailed description hereinafter, in the system of the invention one end of the transformer primary is connected to the power supply, and the voltage on the other end is connected to a sense input of the microcontroller (preferably a comparator input) by a potential divider. The values of this potential divider are chosen such that the mean voltage at the mid point corresponds to the threshold at the microcontroller input. In this way, the microcontroller can be used to sense the phase of the oscillation in the resonant circuit with a known, small and reasonably invariant phase shift. The microcontroller can measure this phase by recording the time at which the sense input crosses the input threshold in both a positive and negative direction.
Using the system of the invention, during operation there is actually sensed the resonant frequency of the relevant circuit, and thus the display age of the element(s) in the circuit can be calculated from that and from a knowledge of the load area. With this information, complete and accurate ageing correction can be performed.
The driving system provides voltage pulses for the display elements. The required driving pulse size can be calculated from a knowledge of the load area, the display age, and the required brightness; it follows that display brightness can accurately be controlled.
The on/off state of the EL display segments is controlled by switches, and these are most conveniently low cost bipolar (NPN) transistors (FETs could be used but are more expensive). This is counter intuitive, as normally a triac would be used to switch an AC signal. However, NPN transistors will always conduct in the reverse direction, so that "off" segments will see a DC voltage of the same amplitude as the AC voltage applied to the "on"
segments. Well manufactured EL displays can tolerate this DC voltage without adverse effect.
In the invention the state of the display's EL segments can be changed in synchrony with the resonant circuit the microcontroller knows the state of each resonant circuit, and so can change the state of display segments in synchrony with this, enabling grey scale brightness control.
According to a second aspect of the invention, there is provided an electroluminescent display comprising a drive and control system according to the first aspect of the invention.
An embodiment of the invention is now described, though by way of illustration only, with reference to the accompanying diagrammatic Drawings in which:
Figure 1 shows in part the circuitry of a multi segment display together with its driving system controlled by a microcontroller in accordance with the invention;
Figure 2 shows a start up timing diagram for the system of Figure 1;
Figure 3 shows a segment turn off timing diagram for the system of Figure 1; and
Figure 4 shows an outline flow diagram for the program run by the microcontroller used in the system of Figure 1.
Figure 1 shows a display driving and control circuit including a microcontroller (11) configured to control the power supply of a resonant electroluminescent display (12). For reasons of clarity, only three of the display's segments (as 13) are actually shown actually the segments
themselves are merely indicated by the presence of the corresponding driving electrodes but it is common with this type of circuit to operate displays with as many as 20 segments, and more. The only limiting factor is the practicality of controlling the switching transistors (Q2, Q3 & Q4), which is normally limited by the number of outputs (as 02, 03 & 04) of the microcontroller 1 1.
The operation of the circuit is now explained with additional reference to the Timing Diagrams in Figures 2 & 3, and the Flow Diagram in Figure 4. Voltages are denoted VA, VB corresponding to the voltages at points A and B in Figure 1. The currents in the transformer (generally 14) primary and secondary (having a turns ratio N) are denoted II and 12 respectively.
In order to operate the display, one or more of the segments 13 will be turned on by holding the respective segment output (02, etc) in a high state, thereby supplying a base current to the appropriate control transi stor (Q2 etc).
Power is applied to the resonant circuit using transistor Ql (any electrical switching part could be used for this function FET or others but a bipolar transistor is currently the lowest cost solution). Via output Oa the microcontroller 11 switches point A to a positive voltage (usually the supply voltage), thereby providing a base current through Rl to Ql allowing a larger current to flow in Ql's collector and the primary of the transformer 14. The current (II) that will flow in the t ransformer's primary is dependant on the state of the relevant resonant circuit. During the start up phase (the first few cycles of applied power) the load capacitance (determined by the segments 13 whose control transistors [Q2 etc] are turned on) is rapidly charged via the transformer whilst transistor Ql is turned on, and current is limited only by the base current applied to Ql (limited by the value of Rl). Once oscillation of the resonant circuit reaches the steady state phase, the output voltage at point B of the circuit will be close to N times the supply voltage (N is the transformer turns
ratio) at the time when Q l is turned on, and a smaller amount of current will flow in the primary to boost the output voltage to N times the supply voltage.
Transistor Ql is turned off after applying to the resonant circuit in this manner a short pulse of current (a typical duration might be less than one eighth of an oscillation period, denoted TPW in Figure 2), and the inductor and capacitor resonant circuit is allowed to oscillate. Resistors R5 and R6 are chosen to have values that are high compared to the rest of the resonant circuit, such that during this oscillation period no significant current flows in the primary side of the circuit. The voltage at the primary winding follows that at the secondary winding (with a ratio equal to the inverse of the turns ratio of the transformer), and can therefore be used to sense the state of the resonant circuit; the microcontroller takes that voltage as an input at Iv.
In the configuration shown in Figure 1, one end of the transformer primary is connected to VDD (the supply) so the voltage on the other end (point C) is given by VC = VDD+VB/N, where VB is the voltage across the transformer secondary (point B). N is again the transformer turns ratio, and is selected such that whilst achieving the required volts at point B the voltage at point C does not go below zero volts (as this would cause the drive transistor to conduct in reverse). In the configuration shown, the voltage at point C is connected to a sense input (Iv) of the microcontroller (preferably a comparator input), point D, by a potential divider. The values of this potential divider are chosen such that the mean voltage at point D corresponds to the threshold at the microcontroller input. In this way, the microcontroller can be used to sense the phase of the oscillation in the resonant circuit with a known, small and reasonably invariant phase shift. The microcontroller can measure this phase by recording the time at which the sense input crosses the input threshold in both a positive and negative direction.
During the application of a current pulse using Ql, the point D voltage at sense input Iv will fall close to zero volts. In the period following this it will rise through the input threshold after a time TR, and then fall back below it at a later time TF. The times of these two events can be used to infer the natural frequency of oscillation of the resonant circuit. The time between the two events will be half of a full period of oscillation, so the resonant frequency, fO = 1/2(TF TR).
The microcontroller can now use this information to adjust the period between applying current pulses (denoted TP), to match the resonant frequency of the circuit. This control technique can be used to keep the resonant circuit oscillating at its natural resonant frequency, resulting in maximum drive voltage to the electroluminescent display and highest power efficiency. It is necessary to determine the resonant frequency, as this will change when the capacitive load changes either because different segments have been turned on or off or because the electroluminescent display has aged resulting in reduced capacitance.
The use of bipolar transistors as control transistors should be noted. This provides significant cost advantage over the normal AC switch component; the triac. The bases of these transistors are held to ground in order for them to be switched off, and to VDD to turn them on. This works as expected, as the transformer secondary sources current through the electroluminescent display towards ground (shown as negative current in Figure 2). However, as can be seen from Figure 2 the transformer secondary also attempts to sink current through these control transistors. When this happens, the collector of the transistors in question is pulled below zero volts, ie more negative than the emitter. As the base is connected to ground for an "off" transistor, and a bipolar transistor is effectively a symmetrical device, it operates upside down (base current
flows through the base collector junction), and conducts. This means that it is impossible to turn the transistor off for the negative phase of the cycle.
The result can be seen in Figure 3. When a segment 13 is turned on, its control transistor Q2 etc conducts during both positive and negative halves of the AC cycle. The collector of the control transistor (point F in Figure 1) stays close to 0V. When the. base is connected to ground (in order to turn the segment off) the transistor continues to conduct whilst the voltage at point B falls to its negative peak, but will not conduct as it rises up to its positive peak. This results in segments in the "off" state have a negative voltage across them (shown in Figure 3 as VF VB). This could be considered a disadvantage as it might be thought to reduce the display life. Tests have shown, however, that careful construction of the display can lead to adequate lifetimes in these conditions.
This ability to turn segments on or off in synchrony with the applied current pulses allows the microcontroller 11 to switch a segment 13 on for a controlled number of cycles. This means that it is possible to light different segments at different brightnesses using the time domain techniques described in the Specification of our British Patent Application No: 0315871.4. To implement this, the length of the current pulse is adjusted by the microcontroller to control the voltage developed across the electroluminescent display dependant on the number of segments lit at the time.
This same control of input pulse length can be used to control the input power dependant on the segments required to be lit at the time, enabling this type of circuit to achieve the display lifetime b enefits resulting from the application of constant power described in the Specification of our British Patent Application No: 0317392.9. Additionally, the ability of this circuit to measure the resonant frequency of the display and transformer combination, and from that to derive the capacitance of the display, enables
the microcontroller to derive the age of the display, and then further to adjust the applied power to compensate, as described in the aforementioned Application No. 0317392.9.
As will be apparent from the foregoing description, the present invention at least in its preferred forms incorporates a number of key differences over the systems presently used in the Art.
Firstly, a digital microcontroller or other sequential state machine is u sed to control a resonant circuit. The systems of the Art are all auto resonant; the invention's use of a sequential state machine controller allows for significant control advantages, as outlined below.
Secondly, in the invention the phase of the relevant resonant circuit is inferred from the primary driving coil. In the Art there is used a separate sense coil; the invention's approach substantially simplifies coil design, and therefore saves cost.
Thirdly, using the system of the invention the display age can be calculated from the resonant frequency and knowledge of the load area. The Art discusses automatic compensation of ageing; as capacitance reduces due to ageing, the drive strength increases. The invention provides the ability actually to sense resonant frequency, allowing there to be calculated the display age. With this information complete ageing correction can be performed, not just some compensation.
Fourthly, the required driving pulse size is calculated from a knowledge of the load area, the display age, and the required brightness. The invention's ability to control pulse width from derived age information and known lit display area means that display brightness can accurately be controlled.
Fifthly, the state of the EL display segments is controlled using low cost NPN transistors. In the invention the display segments are turned on and off using bipolar (NPN) transistors (FETs could be used but are more expensive). This is counterintuitive, as one would expect a triac to be used to switch an AC signal. The recognition that the NPN transistors will always conduct in the reverse direction leads to an understanding that this will work, and that "off" segments will see a DC voltage of the same amplitude as the AC voltage applied to the "on" segments. Well manufactured EL displays can tolerate this DC voltage without adverse effect.
Finally, the state of the display's EL segments is changed in synchrony with the resonant circuit. In the invention, the microcontroller knows the state of the resonant circuit, and so can change the state of display segments in synchrony with this, enabling grey scale brightness control.