Solar buck converter

Author Message

Joined: 09/08/2007
Location: Australia
Posts: 3755
Posted: 08:43pm 17 Nov 2020      

  nickskethisniks said  
In the mean time I've learned more about inductors. So I think I will make new ones with iron powder cores, I plan to stack 2pcs T200-26d per converter. Allthough the ones in the picture will work perfectly on short term.

Micrometals used to have some excellent free on line design software (in DOS!) for powdered iron toroids.  They have since moved everything to China, and all the part numbers have now changed since the old days.
There is some new design software out now, that you need to register for, which I have not yet tried.

Anyhow, for designing inductors or transformers the current and voltage are pure ac with no dc component. As you will already know, once you start getting up into the higher frequencies at high current, skin effect can produce some serious copper losses, and the high ac flux swing in the core can cause unacceptable core heating.

Choke design is very different.
Suppose you have 50 amps of dc with 5 amps rms of superimposed ripple current. As you also know, it will need to have enough copper to carry the dc, maybe 4 amps per mm sq, say roughly 12 mm sq.

Now the interesting thing is that a big fat solid wire like 12mm sq will easily have enough external surface area to carry the 5 amps of ripple without resorting to multi stranded litz wire. The ac ripple will still run just at the surface of the wire with very limited skin depth, but the bulk dc will use the whole wire cross section.

Provided there is enough inductance to limit the ripple component, normal stranded wire will work fine in a buck regulator choke. The pwm inverter guys are using ordinary heavy battery cable in their inverter chokes which works perfectly well.

Its the same with the core losses. Provided the ac flux swing is kept low, you can often get away with low grade powdered iron or even a laminated steel core in a buck converter.
Use plenty of turns !!  
That decreases the ac flux swing (ac volts per turn) but also increases the dc magnetization (ampere turns).

Its all interesting stuff, but an optimally designed powdered iron choke will always use a very large number of turns of solid wire, compared to  gapped ferrite which may require litz or foil.