SMPS converter using NCP1013-130kHz

Thread Starter


Joined May 26, 2014

I tried to build a SMPS converter with the NCP1013-130kHz using a Wurth transformer 7491181112 (which indicated it was intended for the NCP1013-130kHz). It does not have an auxiliary winding, but charges a capacitor on its power supply with an internal "current channel". I have used buck and boost, sepic controller earlier, but never SMPS, so a simply copied the reference design in the NCP1013 datasheet as a starting point. Once done I connected it to the 230VAC outlet and got a very "weak" and pulsing output. I expected 12V but got some 3V (and pulsing). When I unplugged it from the wall it took some 1s and then I got a nice stable 12V for about 1s and then it died. In the data sheet they mentioned a mechanism to detect overcurrent which would shutdown the controller and then try to restart. I connected the converter to 180VDC and then in worked fine. I get the impression that the inductance in the primary winding is to low (it is 2.8mH). Given U/L = dI/dt I set dt to half the clock period (1/(2*130kHz). U was 308V (230VAC) and I is 2,8mH. That gave me dI = 423mA, which is over the limit for the NCP1013 (350mA is max). At 180V I get 247mA, which is ok. I put a series resistor (4,7kohm) on the power inlet and then in worked at 230VAC too.

I am a bit confused why the reference design did not work even though I used the component values suggested and a transformer where voltage levels (80V-385V), SMPS controller (NCP101x), and switching frequency 130kHz indicated it would work. Maybe they only tried it at 120VAC.

What I am missing here?
I series resistor that large does not seem ideal.