Flyback converter design

i dont know how to control two outputs with a single control loop

Neither does anyone else. Your output has two degrees of freedom, and your control input only one. That's a fundamental problem you can't fix with any amount of clever design. Physics can be annoying like that.

However, you can try to make the two outputs track well. One way is to make each as low impedance as possible. The outputs will track each other better the lower the sum of the two open-loop output impedances are.

You also need to decide what is the most important parameter to control. Must each output always have some minimum voltage? Or, is it more important that some maximum output is never exceeded?

Once you know the important control criteria, you can make a circuit OR or AND them into a single signal that says whether you want more or less overall output. For example, if both outputs need to maintain some minimum voltage, but a little higher is OK, then create a single signal that indicates when both outputs are above their minimum. You then use that with a pulse-on-demand controller to decide when to create pulses, and when to stop. The result is that one output will always be at its regulation threshold, while the other might be a bit higher.

Since this is a flyback converter, you also have to make sure the two output windings match their desired output voltages. Make sure to include the diode drops in deciding what each winding must produce. You want the two windings balanced so that the output energy of each pulse is balanced when the two outputs are at their desired levels.

posted about 3 years ago

CC BY-SA 4.0

Olin Lathrop‭

8059 reputation 18 332 961 469

Try searching for "multiple outputs flyback". Basically, you can't control each, individually, instead you make the voltage feedback as the average between the outputs. Here's a quick'n'dirty test to show what I mean, with a simple voltage mode concoction (don't give it too much thought for the values, they work enough for this example):

R3, R4, and R5 form a combined resistive divider which make the averaged value of the outputs appear at their junction. This is the one that the loop considers to follow. The disadvantage is that whatever perturbation exists on one output, only, will affect the other(s), too: R2 is the load for the 24 V (give or take) jumps from 50 Ω to 150 Ω, and you can see how the 12 V output drops a little. It's still stabilized, but it's the average of the outputs that is monitored, not any output in particular.

posted about 3 years ago

CC BY-SA 4.0

a concerned citizen‭

391 reputation 0 16 39 13