Electronics > Projects, Designs, and Technical Stuff
Controller for an Inverse SEPIC (Zeta) Regulator
(1/1)
Odysseus:
I'm hunting for a switch-mode controller to implement a low-output-ripple inverse-SEPIC (Zeta) regulator, as part of an RF application running on noisy USB power.

The best IC I've found so far is the MIC2193, which has the following desired features:
 - High side P-channel gate driver which enables straight-forward Zeta implementation.
 - Synchronous N-channel gate driver for high efficiency.
 - Current-mode control and slope compensation for fast transient response and good stability.
 - Low pin count
 - Low cost

My second choice at the moment is the LTC1773, which is similar but is a bit more expensive and comes with a couple unnecessary pins/features. Figure 10 in that datasheet is close to what I'm going for.

My concern is they are both rather old parts, and stock levels seem low. I'm hoping someone can recommend similar alternatives that aren't as likely to become EOL. Any input is appreciated.
T3sl4co1l:
Haven't looked but going to guess there isn't much out there for integrated P-ch drivers.  Do consider bootstrap (or charge pump if necessary) N-ch high side switches!

Which, actually... if the output voltage is 6-15V, it should be reasonable to bootstrap power it with a diode from GND...  Would still need a bias resistor to charge it up the first time.

...Too bad no one makes drivers that can go below GND, at least for much time. ::)

Anyway, you're not really saving anything by using Zeta over a regular buck or forward converter; that may be easier to solve.  Can always put on more filtering or shielding, nothing terrifically hard about this with respect to RF application. :)

Tim
SiliconWizard:
The TPS40200 does, and it's usable in a Zeta topology. It's also not expensive. https://www.ti.com/product/TPS40200

It doesn't have an additional NMOS driver though, which both ICs the OP suggested have. The NMOS is not strictly required for a Zeta topology AFAIK. Finding regulators with both P and NMOS drivers will be much more challenging...

As to the LTC1773, it's pretty expensive, but I don't know where the OP saw it was old and low stock? It's marked active on AD's website, and there's stock everywhere - sure the stock is limited as with pretty much any "specialty" part that has probably not a lot of demand. (Eg: 440 @Digikey)


Odysseus:
Yeah, pretty sure I'm guilty of being too picky.
Honestly, given that I can easily simulate the LT1773 and guarantee a working solution on the first try, I'm probably going to use it.


--- Quote --- ...Too bad no one makes drivers that can go below GND, at least for much time. ::)
--- End quote ---
I've previously used the LT8471, which has two uncommitted NPN switches with emitters rated to -60V.  As indicated in this app note, it worked nice for generating a clean bipolar supply from USB. I found it rather inefficient, though.
Navigation
Message Index
There was an error while thanking
Thanking...

Go to full version
Powered by SMFPacks Advanced Attachments Uploader Mod