2
\$\begingroup\$

I aim to design a simple open loop non-synchronous buck converter for a project at my college with the following specifications.

Parameter Value
Input Voltage 10-30 V
Output Voltage 5-9 V
Output Current Max 2.5 A
Switching Frequency 50 kHz

My college has used the IR2110 before, and I intended to use a different driver IC.

This is what I came up with

Gate Driver High and Low Side - IR2101 N-channel MOSFET - IRF530

I developed a simulation in LTSpice, using a pulsed voltage source to drive the high-side NMOS. I tried to match the generator according to the IR2101 datasheet. enter image description here

My reasoning for adding a 75 Ω gate resistor instead of a 10 Ω or a 33 Ω is to limit the gate current to IR2101's maximum source and sink capability. Since assuming the fastest turn-on from the total gate charge and rise time of the IRF530, the current requirement is around 760 mA, but the IR2101 cannot supply such. I calculated the inductance first, allowing for a 25% ripple current, then used a 4.7 uF capacitor to see that the output voltage ripple would be around 0.33 V, which is okay for my requirements now.

The simulations show that the gate current (Ig(M1)) is well within the 130 mA / 270 mA source and sink current limits of the IR2101.

enter image description here

The simulations seem to work well, but there are noticeable spikes in the NMOS and Diode currents.

enter image description here

How can I improve the design before moving on to the schematic capture? I am very new to power electronics and would greatly appreciate the help!

Edit 1: As instructed, I have entered properties for the reactive components.

  1. The Inductor is modelled after 1410460C from Murata Inductor
  2. The output capacitor is modelled after UCS2E470MHD1TO from Nichicon Capacitor

The capacitor has been changed to 47 uF to account for the ripple current, which I measured to be 300mA approx. enter image description here

The current spikes in the MOSFET and the Diode can be seen here;

enter image description here

For the diode, the reverse current reaches -121.15 mA before settling at -150 uA enter image description here

For the FET, the spike goes to 2.37 A before reaching 2.21 A and rising steadily as expected.

With lighter load currents, the spikes increase in magnitude.

enter image description here

\$\endgroup\$
3
  • 2
    \$\begingroup\$ Use real models for reactive components i.e. add their effective series resistances (DCR for the choke, ESR for the output cap). That'll give you more realistic results. but there are noticeable spikes in the NMOS and Diode currents. what spikes? Can you mark or even zoom into them on the oscillograms? \$\endgroup\$
    – Rohat Kılıç
    Commented May 4, 2023 at 7:58
  • \$\begingroup\$ Thank you, @RohatKılıç for the advice! As instructed, I am now using real models. I have also indicated the spikes. \$\endgroup\$
    – omen08776
    Commented May 4, 2023 at 11:15
  • \$\begingroup\$ the current requirement is around 760 mA, but the IR2101 cannot supply such Have you tried? It may well literally not supply it: the max current may be exceeded, or may not be. For a one-off testbench thing, it may be possible to just ignore it. In any case, you got the gate resistor, so it's no big deal to change it as needed. You'll need a differential probe for the scope to measure the current waveform (by measuring differential voltage on the resistor). Some gate drivers act as current sources and just won't provide more current. Some don't and get overstressed. Caveat emptor. \$\endgroup\$
    – Kuba hasn't forgotten Monica
    Commented May 4, 2023 at 13:56

0

Reset to default