Half Bridge SMPS gives more stress to its "local" input capacitors?

[cupoftea]

Do you agree that of all the topologies, the Half Bridge SMPS gives the most ripple current stress to its immediately local input capacitors?

Please allow me to explain…

Consider, if you will, a 750W offline Half Bridge (after a 400Vout PFC) where there are say two 6u8 Film caps in the rail splitting capacitor leg of the Half Bridge……upstream of this, say there is a 220uF electrolytic capacitor bank………now consider a 750W Full Bridge, with the same two 6u8 film capacitors (this time in pllel) near to the bridge, and then the same 220uF electrolytic cap bank upstream of that……Now, the Full Bridge will take a significant portion of its ripple current from the electrolytic capacitor bank, and the two pllel’d 6u8 film caps near to the full bridge will be somewhat spared……however, in the case of the Half Bridge…even though the same ‘lytic cap bank exists upstream, the two 6u8 film caps acting as rail splitters will take the full ripple current of the Half Bridge…as such, the stress on these film capacitors is far greater in the case of the Half Bridge…would you agree?

LTspice if a Half Bridge attached.

 

[FvM]

State-of-the art film capacitors 6.8 uF can carry > 10 Arms, what's your problem with a small 750 W converter?

 

[BradtheRad]

Not sure if it's coincidental but your query sounds like it bears on classic paradoxes regarding capacitors.

One is the classic question about two capacitors, one charged and one at 0V, and how much energy is transferred (or lost, or remains in the system) when the two are suddenly connected in parallel?

And also there is the question, Which contains more joules: a) 1 Farad charged to 2 volts, or b) 2 Farads charged to 1 volt? I recall seeing a calculation claiming it's A.

When I think about what is Ampere stress on a capacitor...
I picture electrons racing across the plates, in one direction or the other. Ordinarily we don't think of Ohm's law being at work there, yet it seems as though there ought to be some parameter concerning resistance and heat and Watts lost, greater or smaller according to Volt levels and Ampere levels. I guess it involves ESR value and frequency.

 

[Easy peasy]

for the full bridge the nearest low ESR/ESL caps will supply the ripple ( at twice the volts )

for the half bridge the caps can be rated at just over half Vcc but the current in each will be similar to the above as they supply approx half each ( due to the total ckt with external electro across the total rail )

does your sim show this ?

 

[cupoftea]

for the full bridge the nearest low ESR/ESL caps will supply the ripple ( at twice the volts )

Thanks, and in the case of the top post , do you agree that in the half bridge, the rail splitters will supply **all** of the ripple current, even if a cap bank exists upstream, with much lower esl/esr than the rail splitter caps?

 

[Easy peasy]

The 2 caps supply the sw current AS does the local rail cap - there are several current paths to consider carefully during the on times and off times - draw the ckt and the possible current paths ....

 

[cupoftea]

Thanks, i have attempted a drawing "of sorts" in the top post attache (sim)
I once tested a 3.2kW Half Bridge running off 6 pulse 3phase rectifier...it had two 12uF rail splitters......which would have supplied the ripple....but no bulk caps upstream of that.....which i woudl have thought woudl have been needed to supply the ripple to the rail splitters/power stage....so i guess the ripple ultimately would have to have been taken from the most downstream cap of the input filter module?....which i doubt was very large capacitance.