Stupid question: I'm playing with buck regulators, I've got the PWM signal and a transistor and a supply etc. Maybe I just don't get the point or I'm missing something but.. how do I figure the values for the inductance and capacitance? All the equations I'm finding are for determining frequency, should I match them to the PWM frequency my controller is putting out? I know this is elementary but if you could at least point me in the right direction...
Thanks!
-Eric A.
The values chosen depend on a number of things, and some good app notes exist to help.
Look at the application notes from Linear tech, TI, National, Maxim and Analog devices for starters.
Cheers
PeteS
Key equations you'll probably use: i = C*dv/dt v=L*di/dt
A couple of possibilities exist: you expect the inductor to conduct continuously, or you expect the inductor current to drop to zero for some portion of the cycle. If you expect it to conduct continuously, note that there are basically two distinct periods: one where the input voltage is switched onto the inductor's input, and one where the inductor flies back and its input is near ground (a diode drop below; or a power mosfet drop below). You should know the period of each, from the frequency, the input voltage and the output voltage. During first period, the current in the inductor will increase, and during the second, it will decrease. You know the average inductor current must equal the load current. You want to pick the inductance such that the current at its minimum will be greater than zero (since you expect continuous conduction. Clearly, you should do this at minimum expected load current. Once you've picked the inductance, note the peak current (at max load, just when the first period ends) and find an inductor that won't saturate at that current. Then knowing the current waveform versus time, find the capacitance value that will give you low enough ripple voltage. (Add another L-C section at the output if you want to further reduce the ripple. If you want very low ripple, it will be practically impossible to do it in a single section because of resistances in traces and in the capacitor.)
As another reply noted, there are lots of ap notes that go into more detail. There are several practical things to consider that go beyond the somewhat naive explanation above, but that explanation is the basis for doing the calcs anyway.
Cheers, Tom
Then there is the added nastiness of right-half-plane-zero (RHP zero) instability in continuous mode. This depends on the duty cycle.
Eric: IOW, if this all sounds like Swahili to you I suggest to get a switcher that comes in a can ;-)
This cannot be true.....
Before you blart, surely you must be sure.
How do you live with yourself?
DNA
Well, right now it's me, me CAD program and me glass of Guinness :-)
Have something to add? Share your thoughts — no account required.
Ask the community — no account required