Electronic – Unwanted oscillations and distortions in buck converter

buckdistortionoscillationoscillator

I've tried today to design a very simple and crude 26V to ~12V buck converter. But after testing it in LTspice I've got some weird and unexplained oscillations and distortions. The oscillations produced on 'Q1E' then distorted the signal of the oscillator.

Here's the picture:pic

The green oscillations had a frequency of about 21MHz which it then "mirrored" to the base of Q1. Also the frequency of the 'OSC' had decreased from expected 640kHz to 410kHz and the duty cycle has increased.

If I increase L1 to about 100uH the oscillations are gone and the circuit works as expected. The 'OSC' and 'Q1E' frequency then becomes around 640kHz with a 50% duty cycle.

Why does this happen? The LC resonant frequency of L1 and C6 is only around 16kHz. Does L1 form a resonant circuit with some other capacitor (Q1-BE,D4)?

Is there a way to get rid of the oscillations without increasing L1 to much? I could add a resistor between 'Q1E' and L1 but that would defeat the whole purpose of buck converter.

Best Answer

It could be that the inductor is resonating not with the filter capacitor but with the parasitic capacitances in the transistor and the diode, it could also be that the ringing is being fed back into the LM393 through R10, you mentioned that the ringing was on the base as well? (the output resistance of those things isn't that great). What happens if you use a second 393 as a buffer between U3 and Q1?

Related Topic