Tweet

Posted on 13 August 2019

Flyback Converter and Snubber Design

Free Bodo's Power Magazines!

 

 

 

Every application has its special needs

The Flyback topology is very common for DC/DC converters. Their advantages are the simple design
and the low component cost. The function principle is simple. Only a few components are needed.
The following article gives some practical ideas to design a Flyback converter.

By Ralf Negele, Negal Engineering GmbH Switzerland

 

Flyback Transformer Design

The Flyback topology is very common for DC/DC converters. Their advantages are the simple design and the low component cost. The function principle is simple. Only a few components are needed. The problems are often in the detail. During the design phase several questions arise: Which conduction mode is needed, continuous or discontinuous mode? Which snubber networks are needed? What is important to reduce conducted and radiated emissions?

Nowadays there are a lot of full integrated circuits for flyback converters available, but only a few of standard Flyback transformers. Every application has its special needs. It is not easy to buy “standard” Flyback transformers. So the power supply designer must calculate its own Flyback transformer to get the best performance in cost, size and efficiency.

At the beginning one has to decide, continuous or discontinuous mode. If the operating mode is continuous, one has to consider the recovery time of the secondary diode. As long as the flux of the transformer is not zero, the secondary diode is conducting. It looks nearly like a short for the primary switch if it turns on during the secondary diode is conducting. This results in higher switching losses and higher radiated and conducted emission due to the high current spikes. The problem is reduced if the secondary diode is replaced by a Schottky diode. If the flux of the transformer is zero when the primary switch is turning on, the turn on current spike is much lower. Thus discontinuous operating mode is more practical. The disadvantage of the discontinuous mode is: To transfer the same power to the secondary side a bigger core is needed. This would increase the cost of the power supply.

Discontinuous Mode

During the turn on time of the primary switch, energy is stored in the primary winding of the Flyback transformer. During the turn of time of the primary switch, this energy is transferred to the secondary output capacitor and load.

Figure 1 shows the typical primary current waveform when operating in the discontinuous mode. The input line current is rising linear with a starting point at zero current. The average current needed for further calculation is simply the current IP divided by 2. The starting point for our calculation is the amount of energy needed by our application. With the knowledge of the primary current waveform, we can write the following two equations to calculate the stored energy in the transformers primary inductance.

Primary current waveform

 Equation 1: Stored energy

Equation 1

The needed power for our power supply is calculated as follows:

Equation 2

Equation 2: Input Power

PIN : Input Power
T : Period of one cycle
T : Duty cycle
VPRI : DC-Input Voltage
fS : switching frequency
LP : Primary transformer inductance

With Equation 1 and Equation 2 the maximum primary inductance needed to store the energy in the discontinuous mode can be calculated. For this calculation one should consider the efficiency of the power supply.

Core selection and number of turns

If an appropriate core is selected, the number of turns to store the energy in the core can be calculated with the knowledge of the AL value (core constant). This value is found in the datasheet. Trickier is the selection of the core. For example, E-Cores are very cheap, but they have a higher leakage inductance which in turn can lead to higher emissions. ETD cores have a lower leakage inductance, but they need more space. If we need a high switching frequency of the Flyback converter, we can use ferrite as core material. Maximum core losses and saturation must be considered. If the switching frequency is low, it is likely that saturation will limit the maximum core flux. If we use a ferrite core, an air gap can solve this problem. The air cap should be in the middle of the core. Otherwise we may have problems with radiated EMI. To select the right core some experience is needed.

Secondary inductance

The secondary inductance is calculated in the same way as the primary inductance, except we use the output voltage and power. In order to meet international safety regulations, the transformer in an off-line power supply must have adequate insulation between the primary and secondary windings. We have to know which isolation voltage and the safety creepage distance is required by the applicable safety regulation.

Snubber Design

Flyback converters are using a fast primary switch. The fast turn on and turn off behaviour of the primary switch produces a high du/dt. This fast voltage transition produces an overshoot and lead to ringing waveforms when the switch turns off. That must be properly suppressed. Without this, semiconductors can fail and conducted and radiated noise levels will be higher than necessary. This can lead to higher cost and time when designing the input EMI filter. The high frequency ringing must be damped using RC or RCD snubber networks. On the primary side a possible location for the RC network is between the drain and source of the primary switch. The more effective the RC network is, the more power is dissipated in the resistor. This can lead to lower standby efficiency which could be a problem in green mode power supplies. The best way is if the ringing and overshoot can be minimized or avoided. The source of the problem is the leakage inductance and stray capacitance that leads to ringing wave forms. Leakage inductance and stray capacitance of the transformer are minimized by splitting the primary winding. Further the PCB must be optimized.

In many cases we can not avoid a RC snubber network. Therefore, it is important to optimize the network to reduce the power dissipation in the resistor.

Designing the RC Snubber network

First measure the ringing frequency. As we can see in Figure 2 the ringing frequency is approximately 7.2MHz with a peak of 550V. Now we can add a capacitor parallel to the drain-source of the primary switch. The value of the parallel capacitance should be adjusted until the ringing frequency is 30% less than when we started, in our case 5MHz. We get a capacitance of 100pF. Now we need a series resistor to damp the ringing waveform. The ringing will be well damped if we use a resistor equal to the impedance of the resonance circuit. A good starting point is to use a resistor equal to the impedance of the capacitor. Finally one should consider the maximum power dissipation in the damping resistor. Figure 3 shows the effectiveness of the snubber circuit.

Ringing waveform without snubber network

Ringing waveform with snubber network

Conclusion

During the design phase of the Flyback power supply many aspects must be considered. A good understanding of magnetic core materials is necessary to design an optimized and well working transformer. Also the Layout process and the construction of the transformer are very important. If one could build a Flyback transformer with low leakage inductance and parasitic capacitance, a lot of money and time can be saved to controlling conducted and radiated emissions.

 

 

VN:F [1.9.17_1161]
Rating: 5.0/6 (2 votes cast)
Flyback Converter and Snubber Design, 5.0 out of 6 based on 2 ratings

This post was written by:

- who has written 791 posts on PowerGuru - Power Electronics Information Portal.


Contact the author

Leave a Response

You must be logged in to post a comment.