I was looking at Micrel AN-53 and I have noticed that in a circuit in the first page (circuit is below), at the output of an 400KHz, 27V, 7A, 180W boost converter, there are two 680 uF, two 10 uF and one 0.1 uF capacitor.
I can understand that 680uF capacitors are the bulk output capacitors and instead of 1360 uF, it is better to use 2x680 uF with the benefits of low ESR and rejection of higher frequencies than the ones rejected using 1360 uF.
I may understand that 10 uF is to smooth out even higher frequencies and they have much lower ESR than electrolytics.
However, I cannot understand why is 100nF capacitor is used? Load will use decoupling cap if needed. In addition to that, shouldn't the 100nF capacitor be placed very close to the supply and the ground to keep the lead inductance low? Won't the filtered signal get nasty on its way to the load? Using it in the output makes no sense but unnecessary and extra PCB area to me.
So, could you please explain the function and the need of 10uF and 100nF capacitors in the output (C3, C4, C10)?