Fellow engineers, I recently designed a PCB for converting 12V to 5V/3A using a very common synchronous Buck chip. During full-load testing, I noticed that the overshoot voltage at the SW (switch node) is quite high, with the ringing peak reaching nearly 18V. I added an RC snubber circuit following online formulas, which did reduce the spike slightly, but the high-frequency noise is still coupling into the ADC reference power supply on the same board. Do you have any good suggestions? This board is driving me crazy.
Bro, first confirm if your measurement method is correct. Are you using that long alligator clip ground wire on the oscilloscope probe? That’s an antenna! Remove the cap from your probe and use the ground spring to measure against the GND near the chip—the ringing you’re seeing is likely much smaller; it’s mostly spatial radiation picked up by the probe.
Stop messing with RC Snubbers—they’re just band-aid solutions that don’t address the root cause and actually add losses. Just post a screenshot of your PCB layout. 90% of excessive SW ringing is due to the input capacitor (CIN) having too large a loop area!
Agree with the previous post. Did you place a 0.1uF MLCC small capacitor on the input side? This small capacitor must, must, must be placed right next to the chip’s VIN and GND pins to supply high-frequency transient current. It’s okay if the large electrolytic capacitor is farther away, but if the small capacitor is placed too far, voltage spikes are guaranteed to occur.
Hey, I’ve dealt with nearly the same problem on a 12V-to-5V 3A synchronous buck design. That severe SW ringing isn’t just a simple spike—RC snubbers only damp low-frequency overshoot, and they can’t suppress the high-frequency common-mode noise that leaks into your ADC’s reference supply.The root cause is almost always parasitic inductance in the power loop and SW trace. Try shrinking the high-current loop (input cap → MOSFET → inductor → output cap) to an absolute minimum, and keep the SW trace short and wide. Also, your RC values are probably mismatched—swap to 22Ω + 220pF (a sweet spot for 3A bucks) instead of generic calculated values. On top of that, split analog and digital ground planes with a single-point connection, and add an RC low-pass filter for the ADC VREF to block high-frequency noise. Layout fixes will help way more than just tweaking the snubber alone.
Ugh I totally get how frustrating this is—switching node ringing ruined one of my ADC circuits not long ago, and a standard RC snubber barely made a dent either!For your 12V→5V/3A buck, stop relying only on the snubber first. Move the ADC’s reference power circuit as far as possible from the buck inductor and SW trace, and don’t let analog and power grounds share vias. I also fixed a similar noise issue by using a smaller capacitor in the snubber (100pF instead of larger values) to target fast high-frequency ringing. Throw a tiny ferrite bead on the ADC VREF line too—it’ll block that noisy high-frequency junk way better than you’d think. You’ve got this, it’s just layout and small component tweaks!