Electrical – Fuzzy 3.3V from MC34063

3.3vbucklayoutmc34063output

Recently I've made simple buck regulator based on MC34063 from TI. I want to use that to power my SMT32 base application. Voltage on input varies from 9-26V. I've put some part on general purpose soldering board and while using DMM I got 3.3V on output (in all input range). But when I probe output with oscilloscope I see a lot of noise I'm attaching schematics that I've used. I'm not using any post LC filter only added two 100nF capacitors on input/output to decouple some hihg f noise but that didn't help. Since this is a prototype or prove of concept layout is not optimized (it's very wide) can that play that significant role?

Ringing on switching node

Best Answer

Layout is not optimized (it's very wide) can that play that significant role?

For a DC-DC, layout is everything. Without a proper ground layout, input and output capacitors will be ineffective, and output voltage will be very noisy.

Without a ground plane, there will be large potential difference between different points labeled "GND" due to high di/dt currents flowing through inductive wires. This makes it impossible to measure anything reliably.

Probe GND is max 7cm and I actually keep it close to probe not to make a loop to pick up noise.

You cannot measure anything in the vicinity of a DC-DC with the long ground clip. Lookup "Probe grounding and noise pickup" in this guide:

https://e2e.ti.com/support/power_management/simple_switcher/w/simple_switcher_wiki/2243.understanding-measuring-and-reducing-output-voltage-ripple

Conclusion: Your results are normal for a protoboard. If you want to build a prototype with realistic performance, do it dead-bug over a solid ground plane.