SMPS Design--Theory + Practice: THE DIAGONAL HALF-BRIDGE FLYBACK CONVERTER

Home | Articles | Forum | Glossary | Books

AMAZON multi-meters discounts AMAZON oscilloscope discounts

1. INTRODUCTION

This converter, also known as the two-transistor converter, is particularly suitable for power field-effect transistor (FET) operation. Hence, FET devices are shown in the example used here, but the same design procedure would apply for transistor operation.

The topology also lends itself to all the previous modes of flyback operation-that is, fixed frequency, variable frequency, and complete or incomplete energy transfer operation.

However, there are cost penalties incurred for the additional power device and its isolated drive.

2. OPERATING PRINCIPLE

In the circuit shown in FIG. 5.1, the high-voltage DC line is switched to the primary of a transformer by two power FET transistors, FT1 and FT2. These switches are driven by the control circuitry such that they will both be either "on" or "off " together. Flyback action takes place during the "off " state, as in the previous flyback examples.

The control, isolation, and drive circuitry will be very similar to that previously used for single-ended flyback converters. A small drive transformer is used to provide the simultaneous but isolated drives to the two FET switches.

It should be noted that the cross-connected diodes D1 and D2 return excess flyback energy to the supply lines and provide hard voltage clamping of FT1 and FT2 at a value of only one diode drop above and below the supply-line voltages. Hence switching devices with a 400-V rating may be confidently used, and this topology lends itself very well to power FETs. Moreover, the energy recovery action of diodes D1 and D2 eliminates the need for an energy recovery winding or excessively large snubbing components. The volt age and current waveforms are shown in FIG. 2.

Because the transformer leakage inductance plays an important role in the action of the circuit, the distributed primary and secondary leakage inductive components have been lumped into effective total inductances LLp and LLs and shown external to the "ideal" transformer for the purpose of this explanation.


FIG. 1 Diagonal half-bridge (two-transistor) single-ended flyback converter using power FET primary switches.

The power section operates as follows: When FT1 and FT2 are on, the supply voltage will be applied across the transformer primary Lp and leakage inductance LLp. The starts of all windings will go positive, and the output rectifier diode D3 will be reverse-biased and cut off; therefore secondary current will not flow during the "on" period and the secondary leakage inductance LLs can be neglected.

During the "on" period current will increase linearly in the transformer primary (see FIG. 2) as defined by the equation:

dlp/dt=Vcc/Lp

Energy of 1/2 Lp 2p will be stored in the coupled magnetic field of the transformer, and energy of 1/2 Lp I2p in the effective leakage inductances.

At the end of the "on" period, FT1 and FT2 will turn off simultaneously, and the primary supply current in the FETs will fall to zero. However, the magnetic field strength cannot change without a corresponding change in the flux density, and by flyback action all voltages on the transformer will reverse. Initially diodes D1 and D2 are brought into conduction, clamping the primary flyback voltage (developed by the primary and leak age inductance) to the supply-line voltage. Since the polarity is reversed on all windings, the secondary emf Vs will also bring the output rectifier diode D3 into conduction, and current Is builds up in the secondary winding, as defined by the secondary leakage inductance LLs.


FIG. 2 Primary and secondary waveforms for diagonal half-bridge flyback converter, showing "recovered" energy (energy returned to the supply).

When the secondary current has built up to a value of nIp, where n is the turns ratio and the energy stored in the primary leakage inductance LLp has been transferred back to the supply line, the energy recovery clamp diodes D1 and D2 will cease conduction, and the primary voltage Vp will fall back to the reflected secondary voltage. At this time the volt age across the primary will be the voltage across C3 (as referred to the primary by normal transformer action). This clamped flyback voltage must by design be less than the supply voltage Vcc; otherwise the flyback energy will all be returned to the supply. However, under normal conditions, in a complete energy transfer system, the remaining energy stored in the transformer magnetic field will be transferred to the output capacitor and load during the remaining "off " period of FT1 and FT2. At the end of the "off " period, a new power cycle will start, and the process continues.

3. USEFUL PROPERTIES

This type of converter has a number of useful properties that should not be overlooked.

First (and particularly important for power FET operation), the voltages on the two power devices cannot exceed the supply voltage by more than two diode drops for any operating condition, provided that fast-action clamping diodes are used for D1 and D2.

This very hard voltage clamping action is ideal for power FET operation, as these devices are particularly vulnerable to overvoltage stress.

Second, any energy stored in primary leakage inductance will be returned to the supply line by D1 and D2 at the beginning of the flyback period and is not lost to the system.

Third, under transient loading conditions, if excessive energy has been stored in the transformer primary during the previous "on" period, this will also be returned to the supply line during the flyback period.

Fourth, compared with the single-ended flyback converter, the power devices may be selected for a much lower operating voltage, since the doubling effect that occurs with a single-ended system is absent in this topology.

Finally, a major advantage of this technique is that a bifilar-wound energy recovery winding will not be required; hence the cost and a possible source of unreliability are eliminated.

4. TRANSFORMER DESIGN

The hard voltage clamping action of the cross-connected primary energy recovery diodes (D1 and D2), and the preference to operate at higher frequency with FET devices, means that the primary and secondary leakage inductances of the transformer will play an important role in the operation of the supply.

The energy stored in the primary leakage inductance LLp cannot be transferred to the output circuit; it gets returned to the supply. Hence the leakage inductance results in a useless (loss-generating) interchange of energy in the primary circuit. Also, the secondary leakage inductance results in a slow buildup of current in the secondary rectifiers during the flyback period. This delay means that an additional proportion of the stored energy is returned to the primary circuit and will not be transferred to the output. This proportion increases if the frequency is increased, and clearly the leakage inductance must be minimized for best performance.

A further basic difference between the performance of this arrangement and that of the normal single-ended flyback converter must be considered in the transformer design. In the single-ended flyback converter, it is common practice to allow the flyback voltage to be as large as possible so as to drive the secondary current more rapidly through the output leak age inductance. In the diagonal half-bridge flyback converter, the flyback voltage cannot exceed the forward voltage, since the same primary winding carries out the forward polarization and reverse flyback energy return functions. Hence, because of the hard clamping provided by the primary diodes D1 and D2, it is not possible to increase the primary flyback voltage above the supply lines, and for this application, it is particularly important to design the transformer for minimum leakage inductance.

When selecting secondary turns, the transferred secondary flyback voltage as applied to the primary should be at least 30% lower than the minimum applied primary voltage; otherwise an excessive proportion of the stored energy will be returned to the input line via D1 and D2 at the beginning of the flyback period.

In all other respects, the transformer design procedure is identical to that in the single ended flyback case, Part 2, Sect. 2, and the same procedure should be adopted.

5. DRIVE CIRCUITRY

To ensure rapid and efficient switching of the power FETs, the drive circuit must be capable of charging and discharging the relatively large gate input capacitance of the FETs quickly.

Special low-resistance drive circuitry should be used for this application.

6 OPERATING FREQUENCY

The use of power FETs permits efficient high-frequency operation of the primary power switches. The size of the transformer and output capacitors may be reduced at high frequencies, but the leakage inductance of the transformer, the ESR of the output capacitors, and fast recovery of the rectifiers now becomes particularly important. Therefore, for high frequency operation, not only must the transformer be correctly designed, but the external components must also be selected correctly.

7. SNUBBER COMPONENTS

Because power FET devices are not subject to the same secondary breakdown mechanisms that occur with bipolar devices, from a reliability standpoint, it is often considered that snubber components are not essential. However, in most FET applications, a small RC snubber network will still be fitted across the FETs to reduce RF radiation and meet the dv/dt limitations of the FET. (With very high dv/dt, some power FETs display a failure mode resulting from conduction of the internal parasitic transistor.) However, it is true that the larger snubber components normally associated with reducing secondary breakdown stress for bipolar transistors are not required with power FETs.

To reduce the length of the primary HF current path, a low-inductance capacitor should be fitted across the supply lines as close as possible to the power switches and energy recovery diodes D1 and D2. This is particularly important in high-frequency converters.

8. QUIZ

1. How does the primary topology of the diagonal half-bridge flyback converter differ from that of a single-ended flyback converter?

2. What is the major advantage of the diagonal half-bridge topology?

3. Why is the diagonal half-bridge flyback converter topology particularly suitable for power FET operation?

4. Why is the leakage inductance in the diagonal half-bridge topology particularly important to its performance?

Top of Page

PREV. | Next | Guide Index | HOME