We already know that single switch isolated topologies such as forward or flyback have become the low-cost preferred solution for medium power applications. The simplified circuit schematic is shown in the figure below.
Recall that this basic topology can be used in either a forward or flyback design, with the only difference being the polarity of the transformer secondary winding and the need for an output inductor and freewheeling diode in the forward design, which are not present in the flyback.
Before making any circuit improvements, it is necessary to understand one of the inherent characteristics of the transformer, namely core reset.
Experienced power supply designers will immediately recognize that the above circuit will not work without resetting the transformer. It is important to understand that when current flows through the transformer windings, a magnetic field is formed in the core, which, if not limited, will eventually cause the core to saturate and cause circuit failure. In addition, the core windings produce a magnetizing inductance, and when the switch that controls the current is turned off, the current continues to flow, causing the voltage to rise until a new (freewheeling) path is found. Using a single switch can only magnetize the core in one direction, and both of these problems (core saturation and inductor voltage rise) must be solved. This can be visualized by the core characteristic curve, as shown in the figure below.
When the switch is turned on, the current in the primary winding increases, causing the flux density to increase along the right edge of the curve. When the switch is turned off, the flux density drops to the left edge, but stops before reaching the origin due to hysteresis caused by core losses. This becomes the starting point for the next power pulse, which drives the flux a little higher. Without a reset mechanism, continuous pulses cause the current to move upward periodically, eventually reaching saturation, which causes the current to reach destructive levels.
There are many ways to solve the saturation problem. The earliest and simplest method is to add an extra winding to the transformer after each power pulse, as shown in the forward topology diagram (below).
All three windings are wound on the same core. The reset winding usually has the same number of turns as the primary winding, but can be made of thinner wire because it only needs to flow the magnetizing current, which is much less than the total current of the primary winding. As can be seen from the schematic diagram, the reset winding is connected in series with the diode. When the polarity of the reset winding is opposite to the diode conduction, the diode will block any current from flowing. This will occur during the period of no switching action and during the period when the switch tube is turned on. However, when the switch tube is turned off, the excitation inductance forces the polarity to reverse and couples to the reset winding, allowing the magnetizing current to return to zero by flowing back to the power supply.
The advantages of this technology are as follows:
-
The reset energy is fed back to the input power supply, and the core is reset;
-
Since the reset current is small, the added winding will not add too much difficulty to the design and manufacture of the transformer, and the size of the transformer will not increase too much. At the same time, the only added component is a small diode.
-
If the primary coil and the reset coil have the same number of turns, the drain voltage of the switch tube will be clamped to twice the input voltage.
However, there are some disadvantages:
-
The reset winding and the primary winding have the same number of turns, and the reset voltage is equal to the power supply voltage, which limits the maximum duty cycle to 50%. In the case of unequal turns ratios, the maximum duty cycle can be extended, but the switch tube needs a higher withstand voltage;
-
The added winding still increases the cost of the transformer (process cost, material cost, etc.);
-
There are transient switching losses during the switching process because the drain voltage increases before the current is transferred to the other transformer winding.
Another way to achieve core reset is to use a clamp circuit consisting of a resistor, capacitor, and diode (RCD), as shown in the figure below.
Although this circuit can be used in both forward and flyback applications, it operates slightly differently. In a forward converter, it provides a path for the magnetizing current to return to zero, while in a flyback converter, it uses a capacitor to clamp the voltage spike caused by the leakage inductance between the primary and secondary windings. In both cases, the energy coupled to the capacitor is dissipated in the resistor. The benefits of this approach are:
-
If the added reset element costs less than the additional windings on the transformer, it is a lower-cost solution;
-
Since the voltage across the capacitor increases as the duty cycle increases (thus reducing the off time available for the capacitor to discharge), the maximum duty cycle can be slightly over 50%.
The key issue is that there is loss in the resistor, which is a passive reset method.
In pursuit of high efficiency, we have a third solution, which is to use a MOSFET as a switchable clamping switch to replace the diode. This can still achieve the reset and clamping functions, but also eliminate the overlap time of most current and voltage waveforms during switching, greatly reducing the switching loss of the switching device. Since the switching MOSFET is connected in series with the clamping capacitor, this method is called "active clamp reset" and can use any of the configurations shown in the figure below.
Both circuits work in the same way. The difference is that N-FETs offer more options in high voltage applications, but require floating ground drive, while P-FETs use ground referenced gate drive, but are generally lower voltage ratings. In both cases, the active clamp switch is similar to the synchronous rectifier in a Buck converter, with an on-time of (1-D), and a dead time with the main power switch when both switches are off during the switch transition.
During the dead time, the transformer's inductor and clamping capacitor (including the parasitic drain and source capacitors in the FET) oscillate to achieve resonant conversion. In this way, before the current flows through the switch, the voltage across it has dropped very low, which is often called "zero voltage switching" or ZVS (in fact, it is not a true zero voltage, but has a forward voltage drop of a body diode).
The following figure describes the detailed explanation of resonant transition, which includes 4 different states in a complete switching cycle.
The first state is when the main power switch (Q 1 ) is on for the entire switching cycle, time “D”. In this state, the clamp switch (Q 2 ) is off and the load current is flowing (as shown by the arrow) through the transformer, inductor and main power switch to ground.
The second state starts with the main power switch off. Now the primary inductor forces the current to flow to ground through the clamp capacitor and the body diode of the clamp switch. With the body diode conducting, the clamp switch can be turned on under ZVS conditions.
The third state starts with the activation of the clamp switch and lasts for a "1-D" time. During this period, the current resonates to zero and then reverses to flow toward the power rail. As the current reverses through the magnetizing inductor, the core is not only reset to zero, but also pushed to operate in the third quadrant of the BH curve. This state ends with the clamp switch turning off.
The fourth state, when both switches are off, is a dead time, but the transformer inductance again forces the current to find another path. However, this time it continues to flow upward through the body diode of the main power switch. When the power switch is activated, a new cycle begins, but the current is reversed and is also turned on under ZVS (or close to ZVS) conditions.
Therefore, by adding a second switch and its corresponding gate drive circuit, a new topology is obtained, which has the following advantages over the traditional single-switch forward design:
-
ZVS switching can minimize switching losses, which can increase converter efficiency by several percentage points;
-
Transformer reset is lossless;
-
In the case of bidirectional drive transformer core, improving core utilization can reduce transformer size, refer to the magnetic characteristic curve;
-
Active negative clamping can extend the maximum duty cycle to more than 50%:
-
Resonant current waveform produces less EMI;
-
The waveform on the secondary side of the transformer is very suitable for using synchronous rectifiers.
In the flyback circuit, this topology can also provide many similar advantages, but it has some usage restrictions because the flyback circuit has only a low magnetizing inductance and part of the resonant current on the primary side will be transferred to the secondary side.
提升卡
变色卡
千斤顶
京公网安备 11010802033920号
