SMPS Oscillating Output

I designed a SMPS and made a PCB for it. It successfully can output 6.4V on the output, but whenever any kind of load is attached to it, it starts oscillating for some reason. The following is the schematic and PCB layout.

A few things I’ve noticed that may be a problem :

  • The voltage at the Vcc pin of the NCP1070 PWM is 7.4-7.5V when the min is 7.8Vaccording to datasheet
  • The reference voltage of the TL431AC is 1.25V instead of the desired 2.5V (this is because I assumed the voltage would be 12V at the output prior of assembling the PCB)

These problems may effect the final output of the circuit, but I don’t think they are the cause of the output oscillating.

Hi @thebutterminecutter ,

What is the frequency of the oscillation, roughly?

One thing you could try is connecting a, say 100uF, capacitor from D7 cathode to AC-ground. Does it make any difference?

Cheers, heke

1 Like

Hi @heke,

No, adding a capacitor does not change anything. Other people suggested that I change the D6 & D7 into a Schottky diode because it’s faster. They also said that R3 is not doing a good job at limiting current to the PWM, which is causing it to shut off and on. Any thoughts?

Hello thebutterminecutter,

Random thought, what if you removed R3 and connected a 9 VDC battery into the circuit to pins 1 (pos) and 3 (neg).

  1. Double check the datasheet to verify 9 VDC is appropriate for VCC.

  2. As always, be safe working on these line powered circuits. The battery will float along with the power line. Don’t touch the battery while the device is active.

  3. Use an isolation transformer plus a ground fault interrupter.

  4. Don’t work alone!

Best Wishes,


P.S. I focus on safety from experience. Things can and will catch fire and electrical shocks are serious.

Hi @thebutterminecutter ,

OK. I wonder, could you tell what output voltage you are after?
The circuit would work for 5V, but if you try to get higher voltage then the Auxiliary voltage is so high that the current through the R3 is too high, turning the switcher OFF. Try increasing the R3.
Cheers, heke


I tried increasing R3 to 1K and nothing changed. I went up to 10K but still nothing changed. I initially wanted it to output 12V, but now realized that the transformer is not built for that (5V). So, now my focus is just to get rid of the oscillation. 9V is to high as a Vcc, the max of the PWM is 8.2V

Voltage divider for the battery?

Yeah I tried but there are still oscillations. I was able to get it working by changing R4 & R5 to 1K each and changing R3 to 10K. It now supplies 4.3V and is able to power a DC motor. Although, I don’t think its able to supply a lot of current since I place four LEDs and the motor in parallel and its starts to oscillate again. Is there a way that I can increase the current it can supply, without adding new components?

Looking back at the datasheet, the MOSFET has an R_{DS(on)} = 4.7 \ \Omega.

The chip will have considerable self-heating if a large load is applied.

I wonder if your oscillation isn’t a thermal shutdown.

Have you performed the power consumption calculations for the motor and LED combination?

It may be possible to add an external MOSFET. However, you may be money and time ahead using a different driver chip or topology.

BTW, in case I didn’t say this. Well done!

I get the impressing that you had a considerable increase in your knowledge and skill.



1 Like

Hi @thebutterminecutter ,

When you changed the R4 and R5 to 1k in order to get 5V output, then the R3 cannot be 10k, as it will drop the supply voltage too much. Try 1k for R3, then the current through it will be 2mA, enough for the switcher, but not overloading the internal shunt regulator.
For 12V, try R4 =3.9k, R5=1k (or anything with ratio close to 3.8:1, your originals should be fine), R3=3k9 to yield approx. 4mA supply current. Does that stop the anomaly?
Cheers, heke

1 Like

I don’t think 12V is achievable with the current transformer I’m using, but I’m happy with the 5V output. I decreased the value of R3 and there seems to be no impact, I think the output mostly depends on the voltage divider of R4 & R5. Anyways, thank you guys for everything, I really appreciate it :grin:.

One last question, do you guys know any resource that I can learn more about power electronics?

If you are still interested in getting 12V out of the NCP1070, this reference design from our library may help…

1 Like