Do you think I can make a smaller inductor than the commercial manufacturers?

[dick_freebird]

For all this resolution noodling, just how good do you
require the temperature tolerance to be? And since
the load is not fast or especially "steppy" what do you
think digital control brings to the party? That being the
main benefit I see from digital control loops, breaking
the classical linkage between fSW, control loop BW and
step response.

Still think that sizing the heater element resistance to
the battery natural voltage and bang-bang temperature
control is the way to go. But this project seems to be
always drifting toward the complex and I don't get
why.

Hey, lemonade-from-lemons thought - if you make the
converter super -inefficient- and vaporize using the waste
heat....

But I guess it's not 4/1 anymore.

 

[FvM]

Still think that sizing the heater element resistance to the battery natural voltage and bang-bang temperature control is the way to go. But this project seems to be always drifting toward the complex and I don't get why.

The non-transparent design concept pursued in this thread makes it difficult to help in any way. Some remarks in the parallel thread https://www.edaboard.com/threads/362554/ suggest that the low heater resistance is set by requiring a certain minimal wire diameter due to mechanical constraints. If so, why the term "buck-boost" converter is repeatedly put-in?

If sizing the heater resistance isn't suitable for some reason, supplying it high frequent low-voltage AC could be another approach.

I am really not sure what I should do, should I open a new thread or should I continue the conversation here?

As expectable, the discussion in this thread is partly repeating ideas in the parallel thread, at the borderline of cross posting in my view. The latest questions about pwm generation are well fitting the title of the "structure of a DC-DC/PWM reg" thread, so instead of starting another parallel thread, they should be discussed there.

 

[David_]

The thing is that the situation has become more complicated and the earlier described situation in which one solution would be to tailor the heating elements resistance to a certain value is now only one of the possible uses.

The other use isn't exactly determined yet but it involves using a entirely different kind of atomizer which uses very different coils somewhere around 1Ω or there about.

And as the situation might change further in the future I think it would be best to make the device to be able to be very adjustable.

I got an idea which for sure sounds as further complications(maybe not) but I just thought that given the fact that my µC can change it's clock settings from within a run-time application then maybe it wouldn't be that hard to measure the battery voltage(which I will have to do anyway) and then adjust the clock each 0,1V drop in battery voltage so that the voltage steps keeps at 0,02V more or less.
I at least think that to be a fun idea, I will have to get into programming the µC again to be able to know if that is even plausible because it was a little time ago that I was working with it and all I can remember is loose memories of some capabilities.

It is also possible that I need to stop thinking ahead too much since then I get my self confused and this is after all software problems which can be managed without having to make new circuit boards.

So perhaps I should forget about these things right now and focus on getting the prototype done so that I can finally begin to prove or disprove theories and other questions.
When the prototype is etched I still have possibly quite a lot of hardware things to test and work out, especially considering my limited experience with DC-DC converters.
It is a big adventure that I don't know how it will turn out, failure is a possibility but I would be very surprised if it ended up in total failure.