I read the article "Current Doubler Topology - Myth and Reality " and I think that there are some inaccuracies in the way the author is treating the subject. Upon my experience :(1) by using this topology there is some increase in the total volume of the *2 chokes used (approx. 25%). (2)the advantage in high current outputs is very important due to much easier transformer secondary construction (3)the avoidance of assymetry between two secondaries (center-tap) which in very low voltages can produce transformer satturation if not used a series capacitor with the primary or current mode control (4) the total ripple injected in the output capacitor is absolutely at the level of standard center tap, with relative change at Vin (D.C) change. Actually, there are relative differencies with duty cycle change, but absolutely (over all the Vin range) the changes are minimal.
I built a 5V/50A power supply with following parameters: Vin=32-72VDC Np:Ns=15:3+3turns(C.Tap); 15:6t(C.Doubler) Lf=1.72uH (C.Tap); 2*3.5uH(C.Doubler) Cout=250uF D.C=0.4 to 0.9 The results for the inductor total max. LI^2 are: 1.Center Tap 1.72*52.5^2=4,740mJ (Ipk=52.5A @ D.C=0.4) 2Current Doubler 2*3.5*30^2=6,300mJ (Ipk=30A @ D.C=0.9) THE VOLUME INREASE WILL BE: (6.3-4.74)/4.74*100=32.9%
