This invention relates to a driver for electroluminescent (EL) lamps.
Electroluminescent (EL) lamps are commonly used as liquid crystal display (LCD) backlights for small displays such as in cell phones, watches, pagers, gauges, and portable music players. EL lamps are basically formed of a top transparent electrode plate, a bottom electrode plate, and a phosphor/dielectric sandwiched between the two plates. The phosphor/dielectric may be a sintered phosphor grain layer overlying a dielectric layer. The phosphor glows when a high AC voltage is applied across the electrodes. The type of phosphor used, the phosphor density, the voltage, the frequency, and other factors determine the color and brightness.
The EL lamp is basically a capacitor whose voltage is determined by the charge on its plates, the size of the plates, the thickness of the dielectric, the type of dielectric used, and other factors. The dv/dt across the plates, which is proportional to the current, controls the brightness. The magnitude of current used to charge the plates determines the speed at which the EL lamp charges to its final operating voltage. Once the EL lamp is charged to its final voltage, the voltage stays relatively constant for a short time, depending on the AC frequency, and then the polarity of voltage across the EL lamp is reversed. It is the normal practice to create minimum rise and fall times of the AC voltage since this maximizes the overall brightness of the EL lamp.
The frequency of the AC voltage is in the audible range and is typically 100-2000 Hz. The peak to peak voltage across the EL lamp is typically 100-400 volts. The high voltage (HV) is typically generated from a very low voltage battery (e.g., 1.5 volt) using a boost circuit comprising an inductor, connected to the power supply voltage, that charges by turning on a switching transistor connected to ground and then discharges through a diode when the switching transistor is turned off. A smoothing capacitor is kept at a relatively constant high voltage by being intermittently charged by the inductor at a certain average current and intermittently discharged by the EL lamp at the same average current. The switching frequency of the HV supply is usually at least double the frequency of the voltage across the EL lamp. The HV supply may use any boost technique.
The H-bridge sequencer 12 first turns on transistors 14 and 17 to apply the full HV supply voltage (node 22) to the VA terminal of the EL lamp 20 at a high current to turn the EL lamp 20 on as quickly as practical to achieve maximum brightness. The high current is achieved by large gate widths of the transistors. After the short rise time, the EL lamp 20 is fully charged to the HV supply voltage. An oscillator then controls the sequencer 12 to turn off transistors 14 and 17 and turn on transistors 15 and 16 to apply the full HV supply voltage to the VB terminal of the EL lamp 20 at a high current.
A short zero voltage interval is represented by the waveform, indicating a non-overlapping conduction interval. The interval may be obtained by turning off both PMOS transistors and turning on both NMOS transistors. This discharges the EL lamp to 0 volts.
Due to the large gate widths, the transistors can conduct relatively large currents while ramping up the voltage to quickly raise the EL lamp 20 to its maximum voltage to achieve maximum brightness. Due to the very fast charging and discharging rates of the EL lamp 20, audible vibrations of the EL lamp 20 are created by the nature of the EL lamp's construction, and the vibrations may be heard as a buzzing by someone close to the backlight.
Techniques to reduce audible noise have been used, such as those described in U.S. Pat. Nos. 6,555,967 and 5,789,870, incorporated herein by reference. In these techniques, the ramping up of the voltage is controlled so that the waveform has an exponential shape Before the voltage ramps up to a maximum, the transistors switch to reverse the voltage polarity, and the waveform then quickly falls in an exponential manner. The half waveform is not symmetrical. As a result, the EL lamp is never fully charged, and its duty cycle is reduced by at least one-third. This limits the maximum brightness of the EL lamp. Further, the rise and fall characteristics are not mirror images, since the fall is abrupt, so the prior art techniques take care of only half of the audible noise problem.
A low noise H-bridge driver for EL lamps is described herein. To reduce audible noise of the EL lamp when driven by an H-bridge, the current through the switching transistors is limited while the voltage across the EL lamp is ramping up or down. This reduces the ramp rate of the voltage across the EL lamp and, as a result, reduces vibrations and audible noise to a lower and possibly inaudible level.
The preferred driver provides a rise time of between 5%-50% of a half period waveform and a substantially mirror image fall time of between 5%-50%. During the middle portion of each switching state, the EL lamp is at approximately a maximum voltage. The resulting half period waveform is substantially symmetrical, and the rising and falling portions of the waveform are substantially linear. If the rise times and fall times are small enough, such as 5%-25% of the waveform's period, the EL lamp will achieve substantially its maximum voltage during the cycle, the audible noise will be virtually eliminated, and the rise and fall times will remain short enough for high EL lamp brightness.
Techniques to limit the current through the transistors include: 1) providing switching transistors with a relatively small gate width and reduced gate source voltage (assuming MOSFETS); 2) providing current mirrors to cause the current through the transistors to be the same as or proportional to a fixed current source; or 3) using a feedback signal to keep the current below a threshold. Other suitable techniques for limiting current may also be used.
Increasing the rise and fall times of the voltage inherently reduces the overall brightness of the EL lamp for a particular frequency. However, for a −6 dB reduction in peak acoustic output (one-fourth the acoustic output), as a result of increasing the rise and fall times, the reduction in overall brightness is surprisingly only about 3%, which would be unnoticeable to the viewer. Much lower peak acoustic output reduction is obtained by further increasing the rise and fall times.
Elements labeled with the same numeral are identical or equivalent.
In the driver 30 of
By limiting the current through the transistors, the waveform of
To obtain a short zero voltage interval, either the PMOS transistors are turned on and the NMOS transistors are turned off, or the PMOS transistors are turned off and the NMOS transistors are turned on, by the sequencer 31 for the interval. The current limiting of the transistors provides a linear ramp to the zero voltage interval state.
The preferred driver provides a rise time of between 5%-50% of a half period waveform and a substantially mirror image fall time of between 5%-50%. During the middle portion of each switching state, the EL lamp is at approximately a maximum voltage. The resulting half period waveform is substantially symmetrical, and the rising and falling portions of the waveform are substantially linear. If the rise times and fall times are small enough, such as 5%-25% of the waveform's period, the EL lamp will achieve substantially its maximum voltage during the cycle, the audible noise will be reduced, and the EL lamp brightness will remain high.
The half period waveform may be roughly trapezoidal with rounded edges. The optimal percentage of the rise and fall times depends on the amount of audible noise to eliminate. In one example where the rise and fall times are each about 50% of the total waveform, the AC waveform will be substantially triangular. However, the rise and fall times will be long, resulting in a relatively low brightness EL lamp.
One way to limit the current through the transistors during the ramping stage of the waveform of
The reduced gate width transistors will almost immediately go into saturation after switching, where the drain-source voltage has a negligible effect on current. In saturation, with a fixed Vgs, the current is limited. When the transistor is saturated during the rise and fall times, and the gate width is sufficiently small, the rise and fall times will be extended beyond the
To further limit the current through the current limited transistors, the drive voltage provided by the sequencer 31 (or other circuit) may be slightly above the threshold voltage. The current through a transistor in the saturated region is approximately proportional to (Vgs−Vth)2.
In one embodiment, the maximum current through the transistors of
In one embodiment, the gate width of the current limited transistors 32-35 is reduced by about 25%-75 % compared to the prior art, assuming all other aspects of the system are the same.
The relative sizes of the transistors 52 and 48 may be set to make the current through the H-bridge transistor any proportion of the fixed current generated by the current source 56.
If it is desired to current limit the high side PMOS transistors 46 and 47, the current mirror of
Alternatively, one PMOS transistor and one NMOS transistor in different current paths may be current limited.
Alternatively, a single current mirror at either the upper common node (connected to the HV supply) or the lower common node (connected to ground) may be used. Either one of the current mirrors in
Other techniques for limiting current may also be used, such as using feedback to compare the current through an H-bridge transistor to a fixed reference and controlling the transistor to conduct a current proportional to the reference.
Although MOSFETs have been shown in the examples, current limited bipolar transistors may also be used.
As the EL lamp ages, its equivalent capacitance decreases and its brightness decreases. With the present invention, as the capacitance decreases, the rise and fall times of the voltage across the EL lamp will also decrease. This decreased charge time will offset the inherent reduction in brightness of the lamp.
Having described the invention in detail, those skilled in the art will appreciate that, given the present disclosure, modifications may be made to the invention without departing from the spirit and inventive concepts described herein. Therefore, it is not intended that the scope of the invention be limited to the specific embodiments illustrated and described.