Buck converter design

[bhoke]

Hi all,

I have a question about buck converter design.

Design a Buck converter that converts 50V into 24V. Suppose that we have a switch that can
carry a maximum of 1.25A and the maximum voltage that it can handle is 65V. The switching
frequency is 40KHz (TS=25µs). Assume that the switch has zero leakage current when it is
OFF, but it has a resistance of 1Ω when it is ON. An available Schottky diode has a voltage
drop of 0.4V when it is carrying 1.25A. Design an inductor that does not saturate with 1.25A
using one of the cores below.
1. Find the duty cycle of the switch as a function of output current in such a way to keep
the output voltage fixed at 24V for continuous and discontinuous current cases. (You
may ignore the resistance of your inductor and the switch for this calculation.)
2. Find the maximum power that the converter can deliver without exceeding the limits of
the switch.
3. Find the minimum value of the output capacitor that causes 0.1V ripple at full load.
First, assume that the output voltage is fixed at 24V. At full load, the converter operates in continuous current mode. The
load current not supplied by the inductor must be provided by the capacitor. From the capacitor current waveform, you
can find the ripple in the output.
4. Find the efficiency of your converter at the maximum power (Here, you need to include
the power dissipated on the switch resistance, the inductor resistance and the Schottky
diode).
5. Make an LTSpice simulation of your design to verify the above findings. Perform the
transient simulation until the steady state is reached. Generate an efficiency report to
check your efficiency estimate.
You may use the SPICE models of MBR360 (Schottky diode) and BUL38D (as a BJT switch) given below. Copy the following
lines into LTSpice as a Spice directive. Insert an “npn” transistor (or diode) into the schematic. Cntl-Right click to see the
attributes of the symbol. Substitute BUL38D (or MBR360) for value. If you prefer, you may also use a “nmos” transistor as
a switch. In that case, you need to supply sufficient gate-to-source voltage to turn ON the switch fully.

Available cores:
Violet toroid: AL=50nH/T2, A=11mm2, Bmax=0.2T, le=37.4mm, OD=14.2mm, ID=8.53mm, H=4.8mm
Micrometals T38-8/90: AL=20nH/T2, A=11.4mm2, Bmax=0.2T, le=21.8mm, µr=30, OD=9.53mm,
ID=4.45mm, H=4.83mm
Micrometals T50-7: AL=4.3nH/T2, A=11.2mm2, Bmax=0.2T, le=31.9mm, µr=9.7, OD=12.7mm,
ID=7.7mm, H=4.83mm
EPCOS B64290-L0044-X830 (N30 core): AL=2200nH/T2, A=12.23mm2, Bmax=0.3T, le=30.09mm,
µr=4300, OD=13.6mm, ID=6.5mm, H=5.95mm

I am not asking for the answer, I am asking for method. I dont want anyone to solve this, I just need assistance.

 

[BradtheRad]

Your inductor must have a Henry value, such that combined with the switching device, they produce an RL time constant which allows 1.25 A to build during 'On'-time.

On-time is less than 1/40,000 second (depending on duty cycle).