I am designing a 1.2kW PFC boost circuit. Here are some of the circuit parameters:

Input Voltage: 220Vrms
Output Voltage : 400Vdc
Output Current : 3A at 1200W
Controller : UCC28019A CCM pfc controller
Total Output Capacitance : 1mF

I have designed an inrush limiting circuit using an NTC and a bypass relay. The relay is controlled by an stm32 microcontoller on another board. When the circuit is powered up the input current intitally goes through the NTC where the current will be limited. After a certain delay the bypass relay closes basically shorting the NTC and allowing the current to flow freely.

The problem Im facing is that I dont really know how to properly calculate the required cold resistance for the ntc. I know that I need to calculate the resistance using the input voltage and the maximum allowed input current but that seems way too simple and Im sure there is more to it. I have provided the schematic below.

 

I know that I need to calculate the resistance using the input voltage and the maximum allowed input current but that seems way too simple

Simple is good. Use the peak voltage of the source for the calculation. Allow some margin for unexpected spikes.

 

Use the peak voltage of the source, and the resistance of the rest of the circuit, assuming the filter capacitors are zero ohms, to see what resistance is required to limit the current to what the rectifiers are rated to handle as an inrush current. AND, also consider that since the NTC limiter is bypassed after the startup, it is OK to also have additional resistance in series that is not NTC.
AND, be sure that the stm32 microcontoller does not default to energizing the bypass relay as it is starting up.

 

Use the peak voltage of the source, and the resistance of the rest of the circuit, assuming the filter capacitors are zero ohms, to see what resistance is required to limit the current to what the rectifiers are rated to handle as an inrush current. AND, also consider that since the NTC limiter is bypassed after the startup, it is OK to also have additional resistance in series that is not NTC.
AND, be sure that the stm32 microcontoller does not default to energizing the bypass relay as it is starting up.

What should the current rating of the NTC be? I calculated that I need about 20 ohms to limit the current to 20A.

 

OK!! If you will be limiting the current to 20 amps, then R=V/I=220/20=11 ohms. BUT I would suggest using 20 ohms so that the resistor can be less expensive. AND ALSO, I suggest a simpler time delay for the short delay time. Five seconds should certainly be plenty of time to charge up all the capacitors that we see. Of course, we do not see any of the capacitors that are creating this inrush current surge.
So please consider for future posts, that people considering an answer to solve a problem really do need to see the whole system, not just the part where the originator sees the results of the problem.

 

OK!! If you will be limiting the current to 20 amps, then R=V/I=220/20=11 ohms. BUT I would suggest using 20 ohms so that the resistor can be less expensive. AND ALSO, I suggest a simpler time delay for the short delay time. Five seconds should certainly be plenty of time to charge up all the capacitors that we see. Of course, we do not see any of the capacitors that are creating this inrush current surge.
So please consider for future posts, that people considering an answer to solve a problem really do need to see the whole system, not just the part where the originator sees the results of the problem.

Thanks for your reply. And yes you are right I need to provide the entire schematic that's my bad. Just one more question that I have is how should I select the current rating for the NTC? The NTC will only experience 20A for a short amount of time and because of the bypass relay the steady state current isn't really there. I can't even find an NTC that is rated for more than 10A.

 

I knew that there was a good reason to limit the current more, to less than ten amps. But zI did not recall why. In HAM RADIO big amplifier power supplies they usually just use a 20 ohm power resistor with the relay contacts shunting it for inrush limiting. That way no NTC device is needed, and it is actually more efficient, because after those few seconds no power is wasted in the NTC resistor, which does need to stay hot to keep the low resistance. So the NTC is mostly used where initial cost, rather than efficiency, is most important.