1
\$\begingroup\$

Edit: A p-channel mosfet to keep DC boost circuit separated from main circuit until its charged worked like a charm.

I'm trying to build a DC-DC boost converter, controlled by a 200kHz PWM signal generated by a PIC16F887 microcontroller to produce a 10V DC out of a 5V USB port as a part of hobby project.

When I seperately supply my PIC and DC-DC boost circuit, they both work well. But when I try to supply it from same line, this happens. Periodically, the power circuit's inrush current causes a voltage drop, which causes a reset. Then it all starts once again until the next reset.

I just read that on some sources it's called a "hiccup fault," but couldn't find a proper way to deal with it. Any suggestions?

circuit irl

sim circuit

(p.s: The load and output current will also be clocked next. The Zeners are to keep the DC-DC boost from over boosting during no-load situation.)

edit: pictures are inlined on request

enter image description here enter image description here enter image description here enter image description here

\$\endgroup\$
9
  • 1
    \$\begingroup\$ Please inline your images. Have you tried just adding more decoupling capacitance to either the boost converter, the MCU, or both? \$\endgroup\$
    – Hearth
    Commented Jul 16, 2020 at 12:49
  • \$\begingroup\$ I'll try few more. Do you have any suggestions on capacitor type? Maybe an aluminum instead of film and ceramic types? \$\endgroup\$
    – glados
    Commented Jul 16, 2020 at 12:57
  • \$\begingroup\$ You need a CLC or perhaps CRC filter between the 5Vin and the 5V for logic, a CLC filter between the 5Vin and the supply to the power converter and you need a better layout that you may not be able to achieve with a breadboard. \$\endgroup\$
    – scorpdaddy
    Commented Jul 16, 2020 at 13:17
  • 2
    \$\begingroup\$ The problem is not PWM but the L-Diode-C path with low ESR \$\endgroup\$
    – D.A.S.
    Commented Jul 16, 2020 at 14:59
  • 1
    \$\begingroup\$ You could add a series diode before your MCU decoupling caps so at least they don't get drained by the boost converter. If the caps can hold out enough during the boost converter surge then the MCU won't reset. As others have mentioned, you should have input decoupling caps on your boost converter as well. There are also fancier PMOS circuits ideal diode circuits with less voltage drop. You CANNOT use just lone PMOS for this task since, unlike a regular diode, this will only protect against reverse polarity and reverse current from a load dump; Additional circuitry is needed for the PMOS. \$\endgroup\$
    – DKNguyen
    Commented Jul 19, 2020 at 5:55

0

Reset to default

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.