Would anyone know how to calculate the surge current of my buck converter.

I supply 45 dc volts and it goes to the 20 amp fuse (tiny) then on to the filters 4 470uf caps in parallel. Most time the fuse will pop, if I bring the dc up from a lower voltage first using autotrans, their is no problem.

This is a ebay buck dps5020. I can measure, but pain, so wondering if someone knows a formula. Thanks

 

Based on the formula for capacitive stored energy E = 0.5 x C x V(squared), within a very short time the fuse is passing an energy of 3.8J.

You could try a time delay/slow-blow fuse, or alternatively wire an NTC in series with the fuse. If the NTC has a cold resistance of around 4Ω, you would be limiting the initial inrush current to a maximum of 12A. You need to ensure that the NTC can pass the maximum continuous current that the circuit will draw.

 

Based on the formula for capacitive stored energy E = 0.5 x C x V(squared), within a very short time the fuse is passing an energy of 3.8J.

You could try a time delay/slow-blow fuse, or alternatively wire an NTC in series with the fuse. If the NTC has a cold resistance of around 4Ω, you would be limiting the initial inrush current to a maximum of 12A. You need to ensure that the NTC can pass the maximum continuous current that the circuit will draw.

Thanks very much, that is what I wanted. Right now using switching to charge caps, then switch in main power. I will check out the NTC

 

The problem with charging capacitors is not new, and it is that at the first instant the effective impedance is ZERO, and so the inrush that first instant is limited by the source resistance, which with a regulated supply approaches zero. Thus a huge inrush current. The traditional solution has been a series resistor to limit the current and then a switch to bypass the series resistance after a few seconds.The switch may be a relay with a short time delay before energising, or simply a manual switch. The NTC device mentioned is another similar method that can work very well, except that it does constantly waste a bit of power staying hot. Not knowing the details of your application makes it difficult to suggest an optimum solution.

 

Thanks, that is how I working it now. Series resistor and manually switching it out. I will automate the procedure later.