Higher switching frequency
Tags: Power supply
Basically higher switching frequency allows smaller magnetic cores to be used with fewer turns. But often you are then constrained by the available winding space for your copper wire. While a toroidal core is the ideal shape for magnetic efficiency, giving the best ratio of Ae/le, E & I cores are far more practical and cheaper to wind. In addition with off-line swichers for example, you have to compromise on best winding arrangements to achieve lowest leakage inductance for safety margin in creepage and clearances. Primary and secondaries are then bobbin wound with deep plastic cheeks between primary and secondaries to provide the necessary insulation barriers to meet a 4kV safety flash test.
Switching loss will go up with frequency as you would expect, requiring faster semiconductors and gate drive circuitry. Then you would probably have to slow down your edges to reduce EMC, trading one against the other.
Hysteresis and eddy current loss in the ferrite will increase with frequency and core flux density as will the ac resistance or skin effect of your wire. This is offset by less turns and less ferrite and working at lower flux density and the use of Litz wire, (if you are very desperate for efficiency).
Unless one is trying to achieve very high efficiency, or you are developing a very high power design, where heat can be a major concern, trying to find the very ultimate optimum is not worth the effort. You only have to look at the many reference designs and application notes from semiconductor suppliers to see what is a popular choice of switching frequency for a particular application.
Switching loss will go up with frequency as you would expect, requiring faster semiconductors and gate drive circuitry. Then you would probably have to slow down your edges to reduce EMC, trading one against the other.
Hysteresis and eddy current loss in the ferrite will increase with frequency and core flux density as will the ac resistance or skin effect of your wire. This is offset by less turns and less ferrite and working at lower flux density and the use of Litz wire, (if you are very desperate for efficiency).
Unless one is trying to achieve very high efficiency, or you are developing a very high power design, where heat can be a major concern, trying to find the very ultimate optimum is not worth the effort. You only have to look at the many reference designs and application notes from semiconductor suppliers to see what is a popular choice of switching frequency for a particular application.
<- - I have an Opinion - ->