Nice H-Bridge driver chip for transformers.

[ZomBiE80]

Hello.
Just a couple weeks ago i had to fabbricobble a new H-Bridge to drive a few transformers in a parallel configuration, but i wanted to keep size reasonably small, excluding MOSFETs. So started searching CCFL driver ic's, but i couldn't find any, i n stock atleast. Next i thought: Screw feedback, i could use such driver for fun too. Only chip i found that would do just what i want is this: https://www.nxp.com/products/power-management/lighting-driver-and-controller-ics/hid-controllers/full-bridge-controller-ic-for-hid-automotive-lighting:UBA2036TS
Perfect!

[T3sl4co1l]

What's wrong with a pair of gate drivers?  Or stepper drivers?  Impossible to say without specs...

Tim

[tpowell1830]

Okay, so you found a gate driver, but this isn't a review... what is it? Can you report how well it works under your project? How does it compare to the many other gate drivers out there for your project?

[Red Squirrel]

I'm actually curious since I've been wanting to mess with this stuff myself.  What is the advantage of an H bridge over just push pull?  Of course with push-pull you need a centre tap on the primary.  Is that really the main disadvantage?

I'm looking at this chip myself for general purpose push-pull applications for transformers: 

https://www.digikey.ca/scripts/DkSearch/dksus.dll?Detail&itemSeq=252346857&uq=636544999524521127

(Sometimes DK links don't work for some reason as they are cookie bound, if that's the case, search for SG3525ANGOS-ND)

[T3sl4co1l]

Yeah, use the linkify button in the top right...

Oh, 3525, yeah it's got totem pole outputs (as opposed to the '24 that has "uncommitted BJTs"), oscillator, dead time and an error amp.  Handy for building blocks, if not quite ideal for SMPS purposes alone.

There's some older flavors of the same basic function, in part numbers that I forget; the TL494/KA7500 (TI, Fairchild versions, two error amps, BJT output) and TL598 (totem pole), and the upscale UC3525 and such (Unitrode-now-TI made improved versions of a number of parts like this).

As for PP vs H, more power of course.  Simpler winding helps.  Clamped commutation helps.  Eliminates the switching loss due to primary end-to-end leakage inductance.

PP, half and full bridge are the active-switching versions of familiar rectification circuits: FWCT or current doubler, FWB and voltage doubler.  Note that the voltage and current doublers are merely the series-parallel, L-C duals of each other, so are actually the same circuit fundamentally.

Tim

[NiHaoMike]

Modern full bridge switching PSUs often use a "phase shift" controller, the idea being that both sides are constantly driven at 50% duty cycle (makes it much easier to use transformer drive for the high side FETs) but the phase between them is adjusted to change the effective duty cycle seen by the transformer or whatever it's driving.

[PlainName]

fabbricobble

+100

[ZomBiE80]

Yeah. This is a SMPS chip, not a motor driver. My first project with this chip is using transformers from an old LCD-TV, and those are not center-tapped. I'm looking for use these chips also for other projects which motor drivers are not suitable. I just wanted to hint about these chips that should work nicely on full bridge configuration for many loads except motors. I'll post a review later, when my PCB's are done and working. Driving frequency varies, but might be around 50-200kHz.