Internet Explorer is not a supported browser for TI.com. For the best experience, please use a different browser.

VIDEO SERIES

Common mistakes in flyback power supplies and how to fix them

Learn some of the most common mistakes in the design and troubleshooting of flyback power supplies.When you run into a problem in your power-supply design, the odds are that someone else has already solved the same problem on another design. Wouldn’t it be great if you could learn from their mistakes? This five-part video series focuses on some of the most common mistakes in the design and troubleshooting of low-power AC/DC power supplies, specifically focusing on the flyback topology.

Video Player is loading.
Current Time 0:00
Duration 0:00
Loaded: 0%
Stream Type LIVE
Remaining Time 0:00
 
1x
  • Chapters
  • descriptions off, selected
  • captions off, selected

      Presenter(s)

      [THEME TONES]

      Hello. This is part one of a five-part video series covering common mistakes in flyback power supplies and how to fix them. I am Brian King. My co-author, Mike O'Loughlin, and I have put together this material to help you quickly find the answers to some of the most common issues that you may encounter in your low power flyback power converter. In this first video, we'll go over a quick introduction into the flyback power stage and topology. And then we'll look at some common issues that you may run into when you first power up your power supply.

      The flyback is the most commonly used topology for AC to DC isolated power converters. It offers a very cost-effective way to provide a regulated output which is electrically isolated from the input source. The flyback converter can be broken down into several different subsections, each of which provides a different function. Through the course of this five-part video series, we will touch on common issues related to each of these sections.

      The first section is the bridge rectifier. This full bridge of diodes takes the AC input and rectifies it into a non-isolated, high-voltage DC bus which serves as the input to the flyback power stage. The second section, highlighted here, is the EMI filter and bulk input capacitors. The EMI filter prevents noise from the switching power supply from conducting back onto the input AC lines, while the bolt capacitors store energy between the 50-Hertz or 60-Hertz sine wave AC input cycles.

      The third section is a flyback power stage. When the primary FET, QA turns on, energy is stored in the power transformer, T1. And then when QA turns off, that stored energy is delivered to the output through the output diode, DG.

      Energy flow from input to output is controlled by the fourth section-- the PWM controller. By controlling the power flow, the PWM controller keeps the output voltage regulated within a narrow range. In order to regulate the output, the PWM controller needs to know the state of the output voltage. This information is fed back to the PWM controller through a feedback network, which is labeled as Block 5.

      There are two main methods of providing this feedback. One is the primary side regulation. The other method is the secondary side regulation. In primary side regulation, the PWM controller gets this information about the output by monitoring the voltage on the auxiliary winding of the transformer through a resistor divider network.

      In secondary side regulation, there is an air amplifier and reference located on the secondary side which generates an air signal which gets fed back to the primary side controller. But that signal needs to bridge the isolation boundary, and that is done through an optocoupler. Finally, we need to provide power to our PWM controller. This is accomplished through a startup circuit in the auxiliary winding of the transformer, all of which is shown in block 6.

      For our first common mistake, we're looking at a 24-volt, 1-amp flyback powered from a universal AC input. In addition to the primary output of 24 volts, there is a second output that's needed in this application, which is 15 volts and only three milliamps on the primary side. You can see it on the left side of the screen here.

      So rather than generating a whole 'nother output, we're just going to reuse the VDD rail on the primary side to power this extra output, since it's right around that 15 volts. Now, when we apply power to this supply, it doesn't start up. To debug the situation, we've captured some waveforms, which you can see over here on the right.

      So first of all, on channel number 1, shown in yellow, is the output. You can see it's stuck at 0 volts-- not coming up at all. Channel 2 is the drain of our primary FET, shown in pink here. And you can see that that's actually a channel the scope is triggering on, yet there's no events happening. There's absolutely no switching going on.

      And finally, on channel 3, you can see our VDD rail. And you can see it's stuck somewhere around 5 volts. So what's going on here? Why is this power supply not starting?

      The VDD being stuck at 5 volts is the key piece of information needed to debug this issue. Before the controller can start switching, the VDD voltage needs to rise above a minimum startup voltage, which is around 17 volts for this controller. When the input voltage is applied, VDD capacitance is charged by the current through R1.

      Because we have a 3-milliamp load connected in parallel with this capacitor, the VDD voltage can never rise above the 17-volt startup voltage, because this 3-milliamp load is robbing that startup current. So one solution might be to reduce the value of R1-- to provide more current here to charge that capacitance on VDD. That's an acceptable answer, but if you reduce the value of R1 too much, you're going to have too much power dissipation in R1, and you could have some problems meeting standby power requirements.

      A better solution is to separate power for the auxiliary 15-volt output from the startup current path by adding a second diode, D6, from the auxiliary winding of the transformer. After implementing this change, we now have signs of life, but there's still a problem. Looking at the same waveforms as before, we can see that the output is now rising to about 9 volts. But then it shuts down and continuously tries to restart. So what's the problem now?

      Once again, the VDD voltage waveform is the key clue to debugging this issue. Notice that it does rise to the turn-on level of 17 volts. But once the converter starts switching, it quickly drops to around 8 volts. The problem is that there's not enough capacitance on the VDD rail. Once the VDD voltage drops below a minimum turn-off voltage threshold, the controller shuts off and then has to recharge again before it can try to restart. It's stuck in this continuous cycle.

      The solution is easy. All we need to do is increase the amount of capacitance on the VDD node to provide enough hold-up time for the converter to completely start up. In this case, we increased C3 to 6.8 microfarads. Once the output is completely up, biased power to VDD is provided from the auxiliary winding of the transformer through diode D5.

      To wrap up our discussion on startup issues, we've compiled a short list of tips and tricks. The CDD capacitor needs to be properly sized to hold up the VDD rail during startup. The equation is shown here. In order to minimize the amount of CDD capacitance, look for controllers that have a wide turn-on and turn-off hysteresis and low operating current. Other items that can cause problems are improper number of turns on the transformer auxiliary winding, or extra loading on VDD, or perhaps having a large amount of capacitance on the output, which can increase your startup times.

      View series

      Common mistakes in flyback power supplies and how to fix them