0
\$\begingroup\$

I have designed this buck converter which was supposed to lower the input voltage range 13V - 15V to the 5V and should be capable of delivering 1.5A at 5V. It actually works okay with not load, I measure 5.17V, also works with small load. But when I connected actual load, which would draw current of 1.5A - something just happened, inductor started to make noise and the chip burned(although the white smoke did not left the chip).

Is it possible that this load has peak currents which are too big for the output capacitor? I wasn't able to power it from benchtop adjustable power supply which should be capable to supply 2A - although it works for few seconds on it and then restarts(it is Raspberry Pi with external LCD).

The component pick is coming directly from the TI WEBENCH tool. You can see them in schematics. The actual chip used is TPS563219. *9 is Bourns SDR0805-3R3ML inductor. C20 is C3216JB1A476M. C21 & C22 are GRM31CR61E106KA12L.

This is my schematics and board layout(2-layer, 1oz copper, made in Altium Designer). Beneath whole design there is clear copper pour connected to GND, with no other traces. (S1 is here for different reasons, not connected to the switcher design)

What might be wrong? Is it in peak currents that my load draws? Is the layout wrong at some part? Am I using the wrong chip? (I haven't tried connecting output to osciloscope before burning it).

enter image description here enter image description here

\$\endgroup\$
3
  • \$\begingroup\$ You might want to get out your scope and observe waveforms. "Burning" transistors are often a sign of too slow switching and/or/thus too much time spending with too high RDSon \$\endgroup\$ – PlasmaHH Sep 19 '16 at 19:35
  • \$\begingroup\$ How high can transients on your supply voltage get to? \$\endgroup\$ – Andy aka Sep 19 '16 at 20:37
  • \$\begingroup\$ What's the inductance and saturation current for the inductor? \$\endgroup\$ – winny Sep 19 '16 at 22:11
2
\$\begingroup\$

My first thought is the distance C21 and C22 are from U14. The ground loop for C21 and C22 is also unnecessarily large too. Add a capacitor as close as possible to U14 between pins 1 and 3.

The reasoning behind this is that for a buck converter, the full load current is seen as ripple at the input to the converter. Any stray inductance between the converter IC and the input smoothing will cause voltage spikes at the converter input. These voltage spikes will increase as the output current increases, and are worse for higher switching frequencies. The Absolute Maximum voltage input for U14 is only 19V.

You could also try lowering the input voltage to 7V or so and trying that as an experiment, if you can't get a capacitor closer to U14.

Good luck!

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.