0
\$\begingroup\$

I have a project based on the MKE16F256VLH16 MCU. The processor won't run unless the 3V3 and GND are connected to the debugger (a LPC Link2 board), even if the LPC Link2 is unplugged from USB and left floating.

I've measured the 3V3 line (without the debugger connected) and it's 3.45V, another weird thing, since it should be PSU IC is a fixed 3V3 one.

PS: Have checked and U1 is indeed marked N3XF, which means it's the 3V3 fixed output one.

3V3 line with a 10V input into V_RAW Circuit board section containing the PSU and MCU.

\$\endgroup\$
2
\$\begingroup\$

You do not have sufficient supply filter capacitance. Your bypass capacitors handle high frequencies but you need several microfarads to handle low.

That's especially true as you have a switching power supply: the data sheet of your TPS5060430 shows it typically used with a capacitor in the range of 10 - 22 microfarads. Your 100 pF is probably meant as an additional low-inductance capacitor to limit generated interference, not as the main switcher capacitor. Essentially you are trying to run your switcher without a capacitor, and as the scope shows it is quite unstable.

Adding the debugger as a load adds its supply or target VCC detect filter capacitance.

Fortunately you can probably retrofit this; the geometric path to low frequency capacitors is less critical, but try to get some of it close to the switcher. Maybe you can put one side of the capacitor on the same pad as the inductor, and scrape some soldermask off the ground plane nearby for the other side. Or maybe you can put it next to your 100 pF cap and tap it in with little bits of solid wire.

Next time around study the reference design for the switcher, not just the recommended components, but look at the proposed geometric layout, too.

\$\endgroup\$
1
  • \$\begingroup\$ Thank you! What threw me off was that TI's workbench designer recommended that exact circuit. In any case, another learning opportunity. However, adding a 22uF electrolytic capacitor improved things, but didn't make them as good as I'd have expected. I'm still seeing electrical noise of about 0.4V in amplitude. The average voltage is 3V3 however. \$\endgroup\$ – Alex Nov 7 '20 at 6:39

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.