2

I designed a constant current buck DC-DC super-cap charger with the NCP3420 as a dual MOSFET driver. The current is hysteresis controlled and the disable and PWM lines are pulled low on power-up.

The circuit is working, but sometimes the driver fails during power-up with the controls pulled low, shorting out the 12V supply. What would cause this?

Link to NCP3420 Datasheet: https://www.onsemi.com/pub/Collateral/NCP3420-D.PDF

X1-1 and X1-2 are 12V inputs from my lab supply

X2-1 and X2-2 are connected to a super-capacitor

The 100n cap is a ceramic 50V

The input is also decoupled by 4 parallel 100n ceramics + 6800µ electrolytic

The diode is a schottky rated 80V, 500mA

JG97
  • 124
  • 1
  • 1
  • 7
  • How is the failure fixed? – Andy aka Jun 16 '18 at 10:02
  • I havent been able to fix it yet. The chip needs to be replaced after the faliure – JG97 Jun 16 '18 at 10:05
  • Maybe think about a soft start by adding a capacitor from OD to ground and a 1k resistor feeding that pin from where it is fed from? – Andy aka Jun 16 '18 at 10:12
  • The pin is fed by a pic micro and shold be in high-Z during startup and go to low when the micro starts. Also a valid switch state should not destroy the driver, but the mosfets, since the driver supplies no continuous current – JG97 Jun 16 '18 at 10:17

1 Answers1

0

Think about the Huge low ESR cap that you are charging .What limits the peak current in Q2 ? Also think about the discharge current peaks in Q1 .Is your hysteric current sensing fast enough on R1 ?

Autistic
  • 14,235
  • 2
  • 27
  • 65