Boost and buck converter, preferably with single IC?

[Artlav]

What kind of SMPS topology can produce output that can be either above or below input voltage?
What controller ICs can drive such a converter?

What i want to get is 12V, 200mA output from a source that vary between about 8V and 16V, at high(-ish) efficiency.

 

[skamthey]

You can use Buck-Boost Regulator.!!

 

[FvM]

E.g. Linear LTC3129

 

[Artlav]

Nice, that's about what i'm looking for.
Is there any reliable way to search for them? LTC* is unobtanium here.
"buck-boost" mostly gives inverting or "buck or boost" converters.

 

[mtwieg]

That LTC part is a specialized topology, not your standard buckboost. You're unlikely to find any other ICs like that. A SEPIC topology is probably your best bet if you don't want polarity inversion. SEPIC can work with generic controller ICs. Also you could consider various transformer-based topologies (flyback, push pull, etc).

 

[FvM]

That LTC part is a specialized topology, not your standard buckboost. You're unlikely to find any other ICs like that.

It's synchronous buck-boost topology, a straightforward way to up/down-convert with high efficiency. ICs are available from other manufacturers as well, but Linear has the largest portfolio.

 

[mtwieg]

It's synchronous buck-boost topology, a straightforward way to up/down-convert with high efficiency. ICs are available from other manufacturers as well, but Linear has the largest portfolio.

Buckboost typically refers to the inverting type, with switches in line with both output and input. The synchronous version of a typical buckboost still can only invert the input polarity. What LT is calling a buck boost is something different, where there are four switches which at any given time function as a buck converter or a boost converter, but never as a buckboost. It's really a completely different circuit, and I wish they'd come up with a new name for it rather than improperly using buckboost... maybe bridge converter or something like that...

I'm not aware of any other manufacturer making an IC for controlling that specialized topology, especially one with internal switches. I might not be looking enough though.

 

[FvM]

What LT is calling a buck boost is something different, where there are four switches which at any given time function as a buck converter or a boost converter, but never as a buckboost.

"Never" means, it can't invert the input voltage?

E.g. Analog Devices has this kind of synchronous buck-boost (ADP250x). Also various TI TPS630xx devices.

 

[mtwieg]

"Never" means, it can't invert the input voltage?

Well that's one way of putting it. I put it in terms of the canonical buckboost mechanism: power absorbed only during the on state, power delivered only during the off state. That's what defines a buckboost (or at least has for a few decades), and those ICs can never implement that behavior. Actually, the basic H bridge topology can, in a way, function as a non-inverting buckboost, but only by complementary switching of both sides of the H bridge at the same time (which none of these ICs do, but you could probably do with a standard control IC. Actually I'm surprised I've not seen that marketed before....).

Call me a semantic purist, but topologies have always been classified based on how energy transfer occurs, not just what kind of Vin/Vout ranges are achievable. This new topology isn't a buckboost for the same reason a SEPIC converter isn't a buckboost.

 

[FvM]

power absorbed only during the on state, power delivered only during the off state.

The usual technical term for this is flyback converter, I think.

I have no ambitions either to fight or defend purity in this regard. I just notice that most manufacturers are using buck-boost as a functional rather then topological term. In this sense, it includes SEPIC, too.

P.S.: Technically, these ICs are implementing two coupled half-bridges with different bus voltages. Although the operation principle seems obvious to anyone familiar with synchronous switching in power electronics, the chips aren't available longer than 1 are 2 years. Apparently due to a increasing demand for efficient power conversion in battery powered systems.

 

[mtwieg]

The usual technical term for this is flyback converter, I think.

A flyback is a buckboost-derived converter. All the governing math is exactly the same, except with a turns ratio thrown in and maybe a switched sign convention.

I have no ambitions either to fight or defend purity in this regard. I just notice that most manufacturers are using buck-boost as a functional rather then topological term. In this sense, it includes SEPIC, too.

Yeah, and that bothers me a bit because they're often not really interchangeable, in theory or in practice... oh well, it just means we have to pay attention to what's really going on.