1
\$\begingroup\$

I have an MCU connected to a logic level P-FET(part no AO3401A). The source is wired to a 3.3V regulator that also feeds the MCU. The drain is wired to an LED and a header that connects to a second board with some LEDs and buttons. When I leave the header unplugged the circuit works as expected, and I'm able to turn the LED on and off from the microcontroller. However when I connect the second board the microcontroller browns out and resets continuously. I don't see any fluctuations on the +3.3 when looking at it with a scope.

Why exactly am I seeing these brownouts, and what might I do to prevent them?

Thanks! Circuit diagram below

Circuit diagram.

\$\endgroup\$
  • \$\begingroup\$ The other board may have low ESR caps that induce current spikes which may affect voltage on high impedance Reset. Check or add small cap to it. But it certainly could pose an EMI problem with caps on the other board. Decouple with a PTC or ICL or add a soft start to gate with series R and load C to reduce to 20us ~100us or so. which may cause heat rise. Avoid long unshielded or untwisted power pairs. \$\endgroup\$ – Sunnyskyguy EE75 Mar 13 '18 at 3:25
  • 1
    \$\begingroup\$ 98% that your problem is a mixture of missing/misplaced/inappropriate power supply bypass capacitors on you main board. If this is on a breadboard like this: adafruit.com/product/64 , then you've already found the reason. \$\endgroup\$ – Laszlo Valko Mar 13 '18 at 3:29
  • 1
    \$\begingroup\$ Measure the voltages on all nodes of interest; preferably in the time domain with a storage scope triggered on the gate signal. \$\endgroup\$ – Chris Stratton Mar 13 '18 at 3:49
  • 1
    \$\begingroup\$ Both boards are custom PCBs, connected with an fpc cable. The second board has a 10 uF ceramic cap near the header. How do I determine if this is a problem? I hadn't seen any other power supply issues on the main board until this latest revision where the transistor was added. Thanks everyone for the advice so far, I'll try adding gate resistance/capacitance next \$\endgroup\$ – user181052 Mar 13 '18 at 6:13
  • 1
    \$\begingroup\$ There should be a comparable, dedicated cap on the power supply line, from where the MOSFET is getting fed. As close to the MOSFET as physically possible. In your case, this should be a 10 or 22uF ceramic cap. The trace from the bulk power supply cap to your microcontroller should have as little common section with the trace from the bulk power supply cap to the MOSFET, as possible (preferably zero). If the steady-state current consumption of your other board is rather small, you may introduce a series resistance into the supply line of your other board to artifically limit the inrush current. \$\endgroup\$ – Laszlo Valko Mar 13 '18 at 10:23
1
\$\begingroup\$

I tried adding a 1K resistor to the gate and adding 2.2uF (all I had on hand) capacitor near the transistor but the issue persisted. However when I removed the 10uF cap on the second board everything started working.

The large value was a holdover from a previous version of the design where the second board had a lot more components on it. I'll either downsize it or remove it completely going forward.

\$\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.