Primary side regulated power supplies control the current in the primary side of a multi-winding inductor or transformer to regulate the power Output on the secondary side of the transformer. Flyback power supplies with primary side regulation (PSR) are a type of isolated power supply and are used in many AC to DC applications. For example, many wall chargers use PSR flyback power supplies to charge electronic devices such as cell phones, smart phones and tablet computers. Flyback power supplies control the current through the primary side of a transformer to control the output power. In some embodiments, a controller in a flyback power supply opens and closes a switch that is coupled in series with the primary side of the transformer, wherein the switching frequency is proportional to the output power on the secondary side of the transformer. High frequency switching is used for higher power applications, such as charging applications. Low frequency switching is used when the flyback power supply is in a stand-by mode or where little power is required to be supplied, such as when no devices are being charged by the flyback power supply. In other embodiments, different methods are used to change the output power of the flyback power supply.
In order to achieve very low output and input power during periods when there is no load coupled to the output or the load is drawing very little current, peak current in the primary side of the transformer is reduced. In some embodiments, this is achieved by lowering the switching frequency of the switch that controls the current through the primary side of the transformer. When the power supply is operating in these light or zero load conditions, the power supply is sometimes referred to as operating in a stand-by, sleep, or no-load mode. For example, the switching frequency may be as low as 25 Hz when the power supply is in the stand-by mode. When a device is connected to the output of the power supply, the device draws current and decreases the output voltage. If a device is connected to the power supply between switching states, when the power supply is in stand-by mode, the power supply may not be able to commence the higher frequency switching before the output voltage drops very low. In some situations, the output voltage drops low enough to adversely affect the device that is connected to the power supply.
The input 102 is coupled to the primary side 106 of a transformer T1. In some embodiments, an inductor is used in place of the transformer T1. The primary side 106 of the transformer T1 is coupled to a ground 110 by way of a switch Q1 and a current sensing resistor RCS. In the embodiment of
The output 104 of the power supply 100 is coupled to the secondary side 120 of the transformer T1. In the embodiment of
In the embodiment of
If the output voltage VOUT droops to a predetermined value relative to the last sampled output voltage, the wake node is driven as an output via the auxiliary winding 112. In some embodiments, the wake node is driven as an output when the output voltage VOUT droops to 97% of the last sampled output voltage. When the wake node is an output, the monitor 124 injects current into the secondary side 120 of the transformer T1 until a power cycle from the primary side 106 is detected. More specifically, when the output voltage VOUT droops, a load has been coupled to the output 104 or increased power is being drawn, so the monitor 124 sends a signal via the transformer T1 to the controller 116 indicating the second power level is needed to drive the device coupled to the output 104. In some embodiments, the wake node injects current into the transformer T1 for 1 μs at a 33 kHz rate until the power cycle is detected.
The diode D1 prevents the current generated by the monitor 124 from interfering with the output voltage VOUT. In addition, the wake signal occurs when the diode D1 is reverse biased. The diode D1 also serves other purposes, such as allowing demagnetization current from the transformer T1 to flow into the output circuit following the turn-off of the FET Q1 and then to isolate the secondary side 120 during the remainder of the switching cycle. The diode D1 can be replaced by other devices that perform similar functions.
Having described the basic components of the power supply 100, its operation will now he summarily described followed by a detailed description of the monitor 124. The power supply 100 generates the output voltage VOUT using the input voltage VIN. The input 102 is coupled to the primary side 106 of the transformer T1 wherein current flow through the transformer T1 is controlled by the switch Q1. In the embodiment of
The current flow through the primary side 106 of the transformer T1 determines the output voltage VOUT and the power that the power supply 100 can output. When a load is coupled to the output 104 or additional power is drawn, the controller 116 causes the current flow through the primary side 106 of the transformer T1 to increase enough to supply the second power level to the device coupled to the output 104. When no load is coupled to the output 104 or very little power is drawn, the power supply 100 enters a stand-by mode where very little current flows through the primary side 106 of the transformer T1 and very little power is drawn by the power supply 100. In order to reduce the power drawn by the power supply 100, the switching rate of the switch Q1 is typically reduced. In some embodiments, the switching rate is lowered to low frequencies, such as 25 Hz.
The monitor 124 monitors the output voltage VOUT. When a load is coupled to the output 104, the output voltage VOUT droops. The drooping is detected by the monitor 124, which generates a signal by way of the wake node indicating that a load has been coupled to the output 104. The signal is sometimes referred to as a wake-up signal. The monitor 124 generates the signal by inducing a current on the auxiliary winding 112, which is detected by the controller 116. The controller 116 then brings the power supply 100 out of the stand-by mode and into another mode, sometimes referred to as an active mode, wherein the second power level is generated at the output 104. The process of coming out of the stand-by mode is sometimes referred to as waking the power supply 100. As described above, the second power level may be achieved by increasing the peak current flow through the primary side 106 of the transformer T1, increasing the switching frequency, or both.
In some conventional primary side regulated power supplies the minimum switching frequency is limited to a rate that is high enough level so that the potential droop in the output voltage between switching cycles is held to an acceptable level and the monitor 124 is not required. This minimum frequency comes at the expense of higher input power when the power supplies are in stand-by modes. In some other embodiments where standby power is to be further reduced, the switching frequency can be lower, which causes the potential droop to be much worse.
In some of these conventional power supplies a conventional monitor, that monitors the output voltage dropping to a specific voltage before generating a wake-up signal, is used to transition the power supply out of the stand-by mode. As an example, the power supply may generate five volts at its output. When a load is coupled to the output, the output voltage may droop to a detection threshold of 4.8 volts. In some embodiments the conventional monitor generates the wake-up signal at the next switching cycle after detecting the 4.8 volts. A controller detects the wake-up signal and increases the power available at the output. One of the problems with these conventional monitors is that the detection threshold can interfere with the normal regulation tolerance of the output voltage. For example a 5 volt output may only be able to be controlled to levels of 4.75 V to 5.25 V due to normal component parameter tolerances or operating condition variance. In such an embodiment, a specific 4.8 V droop threshold would interfere with normal regulation of these power supplies. Another disadvantage is that the monitor only monitors for droop on a specific nominal voltage. Accordingly, each monitor will only work with a single voltage output, meaning that a monitor for a power supply having five volt output will not work for a power supply having a 28.0 volt output. Furthermore, if a tight droop tolerance needs to be maintained, i.e. something under 5% for the 5 V case described, a programmable, but non-varying droop threshold is not adequate.
The power supply 100 of
The output 210 of the reference generator 206 is coupled to a comparator 220. In the embodiment of
In the embodiment of
The operation of the monitor 124 will now be described. The sample and hold circuit 200 samples the output voltage VOUT and holds it at the output 204. In some embodiments, the sampling frequency of the sample and hold circuit is the same as the switching frequency of the switch Q1,
When the switch Q1 switches, a pulse is induced on the auxiliary winding 112, which is detected at the wake node. In some embodiments, the pulse is detected on the secondary winding 120 of the transformer T1. The output voltage VOUT is set by the switching characteristics, such as the frequency, of the switch Q1. The sampled and scaled voltage is held at the output 210 of the reference generator 206, which is coupled to the inverting input of the comparator 220. The output voltage VOUT is continually compared to the reference voltage output by the reference generator 206 using the comparator 220.
When an increased load is applied to the output 104 of the power supply 100, the output voltage VOUT droops. However, the reference voltage at the output 210 of the reference generator 206 will not change until the next time that it is sampled because it is held by the sample and hold circuit 200. When the output voltage VOUT droops to a predetermined value relative to the reference voltage, which in the embodiments described herein is 97% of the sampled Output voltage VOUT, the voltage at the non-inverting input of the comparator 220 becomes less than the voltage at the inverting input. When this occurs, the voltage at the output 222 of the comparator 220 changes. The voltage change causes the trigger generator 230 to generate a signal that induces a current in the auxiliary winding 112 to indicate that the power supply 100 needs to output enough power to drive the coupled load. In the embodiments described above, this power is referred to as the second power level.
The monitor 124 generates the wake up signal based on the output voltage VOUT falling below the voltage output by the reference generator 206 and not based on the output voltage VOUT falling below a predetermined fixed voltage. Because of the use of the reference voltage, the monitor 124 may be used with virtually any power supply regardless of the output voltage. For example, if the power supply 100 is outputting five volts at the first power level, the monitor 124 will generate a wake up signal when the output voltage VOUT droops to 4.85 volts between samples of the output voltage VOUT. If the power supply 100 is outputting twenty volts as the first power level, the monitor 124 will generate the wake up signal when the output voltage VOUT droops to 19.4 volts between samples of the output voltage VOUT.
An embodiment of the operation of the power supply 100 is described by the flowchart of
In some embodiments, the sample and hold circuit 200 or other circuit samples several output voltages and compares a selection of them to the reference voltage. For example, an average of several sampled output voltages may he used to generate the reference voltage.
Although illustrative embodiments have been shown and described by way of example, a wide range of alternative embodiments is possible within the scope of the foregoing disclosure. The applicant intends that the appended claims be broadly construed to cover such alternative embodiments, except as limited by the prior art.