Can change of load on one winding affect output of other windings (Flyback SMPS)?

[siliconmike]

For a flyback SMPS, if the output [with feedback] is loaded more, the
switching duty cycle goes up.

In this situation, what would happen to voltages on other outputs ?

(They remain constant I guess, but why?? Shouldn't they shoot up?)

Please explain.

Mike

 

[Klaus Kragelund]

siliconmike skrev:

For a flyback SMPS, if the output [with feedback] is loaded more, the
switching duty cycle goes up.

In this situation, what would happen to voltages on other outputs ?

(They remain constant I guess, but why?? Shouldn't they shoot up?)

Your un-regulated output will increase. The way to avoid this is to:

Stack the windings for better coupling (less leakage inductance). Wind
½ primary first, then all the secondaries and wind the last ½ primary
winding last
Stack the output diodes instead of the output windings
Add a bleeder resistor to the un-regulated winding to pre-load the
output
Use a scheme to not regulate 100% on one winding, but use feedback from
the summing of two windings instead

Regards

Klaus

 

[Tim Williams]

For a flyback SMPS, if the output [with feedback] is loaded more, the
switching duty cycle goes up.

In this situation, what would happen to voltages on other outputs ?

(They remain constant I guess, but why?? Shouldn't they shoot up?)

Nope-- the flyback peak has to do with the magnetic field itself. To a
reasonable approximation, this field is shared evenly by all windings, so
when one drags down the voltage, they all drag down in ratio.

Leakage inductance, winding (and wiring) resistance and other effects
contribute to reduce regulation by decoupling the windings from each other.
If you want one winding better regulated than any other, add feedback to its
output (use an op-amp and opto, if isolation is required). Otherwise, the
usual route (which also allows full isolation) is to add an additional
winding specifically to feed back voltage to the control chip.

Tim