[SOLVED] Flyback Topolgy Help: Why use a 1:1:1 Transformer with secondary tied together?

I am using a switcher chip to make a flyback smps.

There are however two different types of circuits in the datasheet. One type quite a few additional components and uses a transformer with a 1:1:1 transformer (just isolating input voltage, not stepping it up or down) but both secondary coils of the transformer are tied together in parallel like below (I modified it slightly to show the node names):

**broken link removed**

Now the other type uses a normal 1:1 transformer and much fewer components like so:

**broken link removed**


Some questions:

For the 1:1:1 circuit, are the 47ohms and 330pF caps snubbers? Is there some other reason for them? I using low voltage, at somewhat low current (5v @ 500mA) so is it really necessary?

And of course the big one, why use a 1:1:1? If I put a 1:1 inplace of that 1:1:1 with the polarity the same (dot to not dot) would it still work? In the other circuit it is still dot to not dot, but only 1:1, so why would you choose a larger transformer?

This chip does not need a third winding for current sensing, it is all done on the primary side. Is this just a sloppy datasheet, or is there something I am missing?

This is my first isolated power supply so I am relying heavily on the datasheet. Magnetics were never my strongpoint.

snubbers might not be necessary for such low current, BUT it depends on the transistor switch if its rated to handle, usual case of
more robust transistor -> more expensive, would it be cheaper to use snubbers and cheaper transistor....IMO though I still use
the snubbers as I feel sad for the switch being jolted with spikes I can otherwise prevent..... might improve longevity ??

as far as the 1:1:1 I guess, if its already manufactured with 2 secondary windings, why not put them in parallel for more current handling ?
at the same time you have the option of using the 2 windings for a +-V isolated ground referenced...

Adding another identical winding should do nothing to the operation except decrease the effective secondary winding resistance (and maybe change the leakage inductance a bit). That's all. The RC in the first figure is indeed an RC snubber, though it's somewhat uncommon to put it on the output side, though theoretically it should work the same, since the transformer will reflect it to the primary side anyways. In fact in the figure it shows the same network on both primary and secondary.

It's obviously a linear tech document, so I bet they explain things pretty well in there. But anyways, adding the extra windings doesn't change much except the current handling capability.

xaccto said:

snubbers might not be necessary for such low current, BUT it depends on the transistor switch if its rated to handle, usual case of
more robust transistor -> more expensive, would it be cheaper to use snubbers and cheaper transistor....IMO though I still use
the snubbers as I feel sad for the switch being jolted with spikes I can otherwise prevent..... might improve longevity ??

Click to expand...

Makes sense. I guess there is really no harm in having the snubbers there. Part cost is not a large concern at this point.

xaccto said:

as far as the 1:1:1 I guess, if its already manufactured with 2 secondary windings, why not put them in parallel for more current handling ?
at the same time you have the option of using the 2 windings for a +-V isolated ground referenced...

Click to expand...

Well, I have seen many with, and many without 2 secondaries. I don't want to put it on there "just because", I wanted to know why. It seems strange that the datasheet would use a 1:1:1 that way, but I guess no harm done.

mtwieg said:

Adding another identical winding should do nothing to the operation except decrease the effective secondary winding resistance (and maybe change the leakage inductance a bit). That's all. The RC in the first figure is indeed an RC snubber, though it's somewhat uncommon to put it on the output side, though theoretically it should work the same, since the transformer will reflect it to the primary side anyways. In fact in the figure it shows the same network on both primary and secondary.

Click to expand...

Thankyou. Very helpful. I had not seen a circuit similar to this before, so I was thinking maybe there were some unwanted transformer gremlins that could appear and hence the need for the dual snubs. It seems as if the manufacturer is being over cautious. Again, with costs being low for those parts, I see no reason not to include it.

mtwieg said:

It's obviously a linear tech document, so I bet they explain things pretty well in there. But anyways, adding the extra windings doesn't change much except the current handling capability.

Click to expand...

Obvious eh? Lol. Well it is a LT document (LT1424), but I thought the overall concept applied to all flyback's that had no need for a sensing third coil. They explain the theory and operation of the chip beautifully, and I get that. These additional components are not mentioned though. I have only just ventured into SMPS a few months ago. This is my first dealings with an isolated supply. I have basically taught myself the theory behind all the topologies so far from LT/LTC datasheets which are super helpful. I can guess at why parts are there and guess for substitutions but asking for a little clarification before making a board is worth looking stupid for a bit.

I had not considered the increased current handling, or effective resistance change but those effects are really minimal on the circuit's operation which was what I hypothesized. So thanks again to both of you.