From my understanding there is some difficult with compensating a buck converter that is using voltage-mode control compared to current mode control due to that zero (or some other sentence that sounds like that).
From what I've been reading current mode controllers do have a voltage feedback loop, but do not control output voltage directly. So I'm rather puzzled if you can actually use it for charging batteries. At first glance I think no, but then I think "well, is is used to voltage regulators so why wouldn't you be able to use it to charge batteries?"
I've got a bunch of ebooks and printed out a bunch of pages to see if actual texts will explain things better than the glossing they tend to do on the web.
Hi again dk,
Yeah that's the boost or something else, the buck is almost like a regular linear regulator, in fact the averaged model is almost exactly like a linear regulator except it's got a switching element of course. The switching element works very closely like a simple gain so there isnt usually too much of a problem with compensation. The boost does have the right hand plane zero as the control switch acts like a negative gain for some time...it's a very strange setup and we're lucky it works at all as boost can be very handy sometimes
About the different chips...
Take a look at the chips with part numbers like TL494 and similar. These are a bit older type chips but they are quite simple in theory and work pretty well. Im pretty sure they make a MOSFET drive version too, i'd have to look up the number but it's similar to that number.
You can make push pull out of those chips or even buck if i remember right. The duty cycle might be limited to 95 percent though, but that's not too bad as we need some dead time anyway,
and the frequency is variable depending on a small capacitor value.
The nice thing is that they have two error amps built in, one you can use for voltage feedback and the other for current feedback...one loop per feedback, and they are already OR'd together.
As i was saying, the voltage feedback measures output voltage and the current feedback measures output current. If the voltage tries to go over the set point (4.2v) the voltage cuts back the pulse width, and if the current tries to go over that set point (say 1amp) the current feedback cuts back the pulse width. It's not hard to understand really. Take a look at those chips and see what you think. Even if you dont use them you might gain some insight.
In any case, i hope i have been able to help in this thread.