Yes, those terms apply here, in the time domain, for pulse waveforms; it's not just for sine wave steady-state.
Calculating those quantities exactly is a bit of a pain, but the short of it is: you're storing some energy in the inductor, then dumping it into the load. This is reactive power. Reactive power (VAR) divided by switching frequency (and there's a factor of 2pi in there) is the energy stored in the reactance, which is also E = 0.5 L Ipk^2. (There'll be some factors of sqrt(2) or 3 or such in there, for a square wave, but the 2pi alone is correct for the sine wave case.)
Power dissipated in the inductor is VAR / Q, if you know the Q factor of the inductor.
Output power equals VAR if the current is discontinuous (i.e., all the inductor's energy is dumped into the load, every cycle, the current falling to zero inbetween). If current is continuous however (if the switch is turning on while there's still current flowing to the output), output power goes up while VAR remains constant (VAR only depends on the change in current or voltage -- the ripple fraction; it's an AC thing only).
(This is handy when you consider a lossy core material that offers a lot of inductance at a low cost: if the power dissipation limit for the inductor is say 1W, and it has a Q of 10, you can get 10 VAR through it; that's only 10W output for a DCM converter, but 50 or 100W or more for a CCM converter with 20% or 10% or even lower ripple fraction, respectively -- assuming you can fit enough wire on the core to handle that much current, and assuming the core retains enough inductance under bias.)
Otherwise, you might as well calculate the losses in the resistive elements (DCR and core loss), or treat it as a black box (fairly necessary for general SPICE models) and wire a power meter circuit around it.
Tim
Home
Help
Search
Login
Register
